info@analyticsindiamag.com +91-95389 89028, +91-96325 33477

Machine Learning Developers Summit 2024

February 1 to 2, 2024 | Bangalore

INDIA’S NO.1 CONFERENCE EXCLUSIVELY FOR MACHINE LEARNING PRACTITIONERS ECOSYSTEM

about

Topics featured include AI/ML, deep learning, big data, data analytics, security and privacy, enterprise solutions, hardware, software, chip manufacturing, wearables, robots, edge computing, data ethics, voice technology innovation, and accessibility.

The majority of Conference sessions are curated by the AIM community.

We are in the process of finalizing the sessions for 2024. Expect more than 50 talks at the summit. Please check back this page again.

When & Where

Thursday to Friday
February 1 to 2, 2024

NIMHANS Convention Center, Bengaluru, India

What to expect

3 Tracks over 2 days –

  • Keynotes (Hall 1)
  • Tech Talks (Hall 2)
  • Paper Presentation (Hall 3)

Besides – mentoring sessions, Hackathon, Awards, Exhibition & a lot more

Schedule from 2023

Expand All +
  • Day 1

    Jan 19, 2023

  • Rajat tells the story off google's journey into an AI-first world! The story talks about chance interaction,  curiosity about the human mind and above all the googleyness spirit that helped Google create differentiated product offerings by being amongst the very first to leverage AI and ML in their products. He also talks about how you can leverage Google's learning in your own organisation to re-create the same magic!
    Audi 1

  • In recent years’ video action recognition has become a hot research topic due to a wide range of applications such as video surveillance and video analytics. Large volumes of training data, which have more than 56,000 video sequences, are required for training a decent action recognition neural network. However, such large-scale video recordings, which are supposed to capture the dynamics of every action category, are not only prohibitively expensive to gather but also impractical. Training samples are few and rare (e.g. when the target action classes are not present in the current publicly available datasets). For the proposed frugal pipeline, 10% of labeled data is good enough for training an initial neural network to predict the labels for the unlabeled data. Initially, we started with only 10% of the data and passed it through the augmentation pipeline. Then, we used that data to train a neural network and used the best neural network weights to perform pseudo-labeling. All our unlabeled data was labeled during our experimentation at the 4th iteration of the data through our pipelines. It increased the model's accuracy by 10\% on a dataset with no augmentation. This approach consists of two pipelines, the first is a data augmentation pipeline, and another is a pseudo labeling pipeline; types of augmentations can be changed according to the use case.
    Audi 3

  • This workshop will be a curtain raiser to the approach and best practices on operationalising AI/ML solutions from research to production .Borrowing best practices synergies from software development automation .We will walk through a Deep learning use case ,creating pipelines to operationalize the workflow thus created in an interactive workshop session .The attendees will get a first-hand understanding of creating a AI workflow , automating the pipeline and different approaches to training and a glimpse of the road ahead. Prerequisites for the Workshop:
    Docker Desktop (Preinstalled by attendees)
    16 GB RAM local system
    Audi 2

  • Data entered into QuickBooks has real financial implications - a single data entry mistake can be expensive for the business requiring significant effort in undoing the impact. Customer research highlighted the following top concerns: Error in entering invoice data, Error in paying employees, Fat finger a bill payment to a new vendor.. We needed a system which will flag errors on key fields while entering transaction data, and nudge the user towards the correct range. AI solutions are now norm for delivering smart product features to enhance customer experience but most solutions are focused on tailored use-cases that need resources from ideation to deployment and subsequent integration, every time we adopt a new use case. Given the scope of multiple use-cases across data input fields in Intuit products, the error-detection system had to be extensible so that it allows developers to build tailored AI models for specific error detection by providing information about the fields and data sources only. FEDS employs novel methodologies for error detection based on Empirical Bayesian Test, Likelihood Ratio Test and Deep Learning approaches. It decouples the efforts of Data Scientists from the underlying implementation layer, which facilitates the product developers to easily adopt across multiple use cases.
    Audi 1

  • In this paper, we describe a novel case study on the application of recommender systems to generate sales leads for merchants in a B2B environment. A user-based collaborative filtering algorithm is used on corporate payments data to generate new client recommendations for the merchants based on the existing clients of similar merchants. Client recommendations or prospects, if generated accurately, can have huge business potential for the merchants by aiding them in expanding their business. Hence the paper describes an application of recommender systems with significant business impact and value. Despite finding lots of use cases in the B2C environment, recommender systems have till date, very limited applications in the B2B arena. By providing a concrete step-by-step account of how to apply this powerful machine learning technique, our paper aims to firstly bring to light the significant business impact the use of recommender systems may have in the B2B space and, secondly, serve as evidence to be referred to for further applications.
    Audi 3

  • Computer vision AI systems trained on medical images such as X-rays, CT scans etc. can be used to identify, classify and locate specific health conditions, like severity of knee-joint-space, location of vertebrae slip etc., with increased speed and accuracy. Leveraging medical imaging AI tools can provide clinical reviewers with images annotated with relevant clinical parameters, which can aid decision-making across the utilization management value chain, driving improved operational efficiency, accuracy and consistency in decision-making and delivery of health care.
    Audi 1

  • Inventory or stocks are the goods held by any business organisation for the purpose of consumption, production, or sale. These are the non-capitalized assets of the organisation. Inventory can be classified into four categories - finished goods, raw materials, work-in-progress inventory and spare parts inventory (also called as plant maintenance, repairs and operations (MRO) inventory). Spare parts inventory is leveraged in the running and maintenance of the plants or the manufacturing units. In general, inventory management is the process of balancing the inventory thereby - having the right amount of inventory in the right place at the right time. Though it looks easy, it is one of the most challenging problems that industries deal with. In this paper, we will discuss what are the different challenges organisations face during inventory management and how we solve them using artificial intelligence (AI). We have built an AI-enabled dynamic inventory management application across all the inventory categories. Dynamic inventory for finished goods application optimises the inventory using stock, order, product, location and historical sales data across the business. The application can accurately forecast demand to get the right level of inventory, in the right locations, to meet service-level targets. Similarly, dynamic inventory for raw materials application applies AI models to inventory, sales, bill of materials, supplier, work-in-progress and finished goods data from across your business – so you can optimise raw material inventory levels to meet demand and reduce costs. The dynamic inventory for spare parts inventory management application applies AI models to plant data like location and capacity, and inventory data like stores, spares and consumables across the business. We predict demand across stores, spares, and consumables on a granular level. Then, we use forecasted demand to optimise inventory across each plant which will save cost while meeting the required service levels. We will also show how businesses can achieve a good ROI by deriving value from our AI-based inventory management applications using some use-case examples.
    Audi 3

  • Google has been widely recognized as the industry leader in AI/ML. Apart from building Google's own planet-scale consumer applications like Gmail, Photots etc. which are powered by AI/ML, Google has also been a strong supporter of and contributor to the open source community. Open source is indeed at the heart of Google—it’s what we’re built on and at the core of everything we do. From sharing datasets to creating platforms like Tensorflow for building AI/ML models and container orchestration systems like Kubernetes that help scale ML, Google's contribution to the open source AI/ML community has been immense. Join this session to learn more about Google's AI/ML contributions to the open source community, that you as ML Developers can take advantage of, to innovate, build and scale, just like we at Google do.
    Audi 1

  • Preparing an efficient and effective staffing roster is a challenging task for retailers. An efficient staffing schedule optimizes payroll cost while meeting desired customer service levels. Doing this while complying with labour laws and organizational policies makes the schedule effective. While compliance with legal requirements is mandatory, evaluating the impact of alternate organizational policies helps formulate a suitable staffing strategy. These policies present themselves in a combination of factors around shift duration, standardized shift start times, number & length of breaks during the day, off-days, etc. In this paper, we describe our evaluation of alternate organizational policies on the payroll costs of a leading luxury retailer in the UK. Each candidate policy was evaluated using Mixed Integer Linear Programming (MILP), following a two-step approach. In the first step, we forecast the number of customers visiting the store at half-hourly intervals and translate it into the expected staff count required on the shop floor. In the second step, we infuse this, together with legal and organizational policy constraints, into the MILP. Our experiments revealed that flexibility in shift start times, and the mandated minimum staff strength for a store, were the primary drivers of payroll cost. More than three shift start times a day allowed the flexibility of having a higher mix of part-timers. This translated into a 10-12% reduction in paid hours. This benefit can then be considered in the context of the overhead of managing a large part-time population and a potential shift in organizational culture. Further, the driver of minimum staff strength highlighted the need for reviewing the store’s physical layout.
    Audi 3

  • In this talk, we will learn Google's way of putting thousands of models in production and see how our MLOps framework can help you achieve more clarity and faster models in production. We also explore Google's Vertex AI platform, which is being built to handle MLOps at google's scale. Finally, the talk will introduce the Google Cloud MLOps ecosystem and products like Vertex AutoML, workbench, Pipelines, Feature Store, Explainable AI, and Model Monitoring.
    Audi 2

  • Digital advertising heavily relies upon cookie-based solutions to identify and target the prospective audience. With third-party cookies being deprecated in the near future, advertisers will have to look at alternate strategies, such as geo contextual for targeting. In this paper, we have proposed cookie-less solutions for audience expansion by suggesting look-alike postal codes (for geo and audience targeting) and domains (for contextual targeting). These solutions could be used as a part of pre-campaign planning and use customer affinity and geo-contextual datasets along with a universal ad feed as historical learning. We have used deep autoencoders and self-organizing maps for low dimensional latent space representation and probabilistic methods for arriving at insights. A penalized similarity function is applied to this latent space to suggest look-alike postal codes and domains. To overcome the challenge of estimating how good our unsupervised learning models are, we also have developed a framework to threshold the number of postal codes/domains suggested and validated the results.
    Audi 3

  • An intelligent combination of Data-Centric and Model-Centric approaches that not just focus on procuring the relevant data and on algorithmic tools and techniques but also focuses on generating information by emphasising on business relevance, responsible AI, sustainable AI and explainable AI, which collectively help to solve the business problem.
    Audi 1

  • In this talk, I will cover the use case of storing and analyzing images of Yoga Poses with BigQuery and creating a classification model using BigQuery ML to label the postures using only SQL and no other form of code.
    Audi 2

  • SOTA computer vision algorithms play a large role in addressing business problems across verticals such as Manufacturing, Retail, Insurance, etc. In this talk by Praveen Jesudhas, learn how advancements related to active learning, synthetic data generation, few-shot learning, etc., can be consolidated efficiently to enable rapid evaluation and deployment of Computer vision-based solutions.
    Audi 1

  • This paper explores the impact of brand trust on commercial consumer behavior and marketing management in the technology industry, specifically in cloud services. In this study, we examine factors influencing consumer loyalty in products and service businesses leading to trust and love and offer constraints-driven recommendations to promote it. Research focuses on the purchasing behavior of organizations buying goods and services for use in the production of other products or services that are rented, sold, or provided to others. Testing the hypothesis and constructing the model begins with exploratory factor analysis and structural equation modeling. Results indicate that the factors mentioned have an impact on consumers' trust in brands and are presented as a visual network graph. In order to measure the relationship, an evolutionary equation is derived utilizing controllable factors such as service history and surveys and uncontrollable factors such as socioeconomics and market factors. As a result, each account in service is assigned a rank-based time-driven risk profiling index.
    Audi 3

  • Retail manufacturers are grappling to predict customer demand for each product at superior forecast accuracy levels with the ability to refresh the forecasts at weekly intervals due to computational expenses. The overall cost of demand forecasting is determined by adding product shortage cost (loss of missed customer demand), excess product cost (inventory holding costs), product pilferage cost (expired goods cost), and computational cost (involved in forecast models training). While the first three cost components can be attributed to the forecast accuracy levels, the computational cost lever can be controlled through computationally light forecasting systems. The motivation of this research is driven by very limited empirical evidence on the adoption of unsupervised techniques to identify groups of products with similar levels, trends, and seasonality patterns for the development of computationally light forecasting systems. This study aims to decrease the forecast model training cycles by leveraging unsupervised techniques like K-means Clustering and Hierarchical Clustering to identify clusters of products with similar customer purchasing behavior. This research introduces a Clustering-based Demand Forecasting Framework that determines the best forecasting algorithm for each cluster. The experimental approach utilized this framework to predict the customer demand for more than 500 dairy products for the next 8 weeks. A comparative study on computational time across product-level model training and cluster-level model training is presented for better realization of relaxation in computational costs. The Weighted Accuracy Percentage Error (WAPE) based product level forecast accuracy from the Clustering-based Demand Forecasting Framework was found to be superior to the Standard Demand Forecasting Framework. Further, this research provides better interpretability of clusters by providing a product category-specific name to each cluster through Natural Language Processing (NLP) techniques. Holistically, this research effort provides empirical evidence to retail manufacturers for enterprise-wide adoption of the computationally light short-term forecasting system.
    Audi 3

  • According to a study by Mckinsey, 71% consumers expect personalisation and 76% get frustrated when they don't find it. As we see a surge in digital behaviour, personalisation has become a necessity for most businesses. However, while most people lay stronger emphasis on the recommendation engines powering personalisation, design and engineering play an equally critical role in delivering maximum value through personalisation, along with data science.
    Audi 1

  • Traditional numerical methods of classification of both univariate and multivariate time series have shown to be challenging due to their problematic features, including the inability to preserve temporal correlation, lack of pre-trained models, difficulty in training Timeseries models and finally, Timeseries tend to act incorrectly when presented with multiple input. Decades have shown immense importance in applying deep learning methods to image and video data to solve real-world issues. The authors are motivated by recent accomplishments of supervised learning approaches in computer vision, so they have explored various time series image encoding techniques to enable machines to visually identify, classify, and learn structures and patterns by leveraging state-of-the-art Deep learning and computer vision. The Authors demonstrated rapid spot groupings of series with specific features. Later, these series were classified and subjected to comparison analysis using multiple cutting-edge deep neural networks to their maximum performance potential. Finally, the comparison results were demonstrated.
    Audi 3

  • Lookalike audience generation is an effective way to increase the audience base in online advertising. Segregating the lookalike audience into multiple priority levels gives greater flexibility to the advertiser in selecting their user reach. This paper explains a novel approach to lookalike audience generation in multiple priority levels on large-scale data with millions of users and thousands of audience segments. An automated system combining custom models to generate similar audience segments and group lookalike audiences into priority levels using Spark Scala and the Hadoop ecosystem is developed. The experimental results comparing different approaches show that our proposed model outperforms others in reach, scalability, and speed.
    Audi 3

  • In information extraction, not only extracting the entities but also identifying the relationships between the entities can lead to better hidden insights. Triples are used to represent the relationships between clinical entities. Triples are made up of two concepts and a relationship between them: . Graph data is prominently used to extract vital hidden insights by node-edge analysis. This research studies different graph embedding algorithms to predict the relationship between nodes. The paper examined the state-of-the-art in clinical relation prediction and created a graph-based model to predict the relationship between two clinical concepts. The relation prediction model is trained using the proprietary Graph dataset, maintaining the concept and relation triples. The work implements and compares the results of link prediction algorithms in the clinical domain. The ComplEx graph-based node embedding algorithm has been found to be promising in predicting semantically enhanced clinical relationships. The paper discussed the potential use cases of clinical relationship prediction in detail. This research will be critical in determining the causal relationships between medical concepts.
    Audi 3

  • As more organizations look to undergo a transformation, it is critical to understand how data and analytics are the thread that carries the project through—start to finish. The stakes are high for any organization coming out of the last few years. As the marketplace faced unforeseen disruption, it became imperative for businesses to digitize how they engage with both employees and customers. Whether the goal is to sustain, compete, or disrupt—organizations of all types, sizes, and industries have had to reimagine their business models and processes. The phrase that’s been on the forefront of every business-centric conversation for a while has been “digital transformation” and for a transformation to be successful, you need to implement the right data strategy, technology, and infrastructure to modernize your company and become a more competitive and agile player in the marketplace. It requires technology, digital processes, and organizational change. It also requires a focus on data and analytics because the two are key accelerants of any transformation effort. As technologies and consumer expectations continue to evolve, businesses need to keep pace. A company’s ability to compete and thrive will be determined by its ability to make quick, informed data- driven decisions. Where data provides the information, analytics provide the insights that lead to informed decision-making. Using data and analytics as the thread that carries your transformation—before starting and throughout—allows you to overcome key modern business problems that can be roadblocks to successful data initiatives.
    Audi 1

  • Temporal disaggregation is a method of disaggregating low frequency time series into higher-frequency time series. A major challenge faced by many when working with time series data is the non-availability of data at the required granularity. One might come across instances where the data is monthly or quarterly, but the preference is for the data to be weekly, daily, etc. Commonly used techniques to estimate the missing values often fail to capture the behavior of the underlying high-frequency data. This paper introduces a novel approach to tackling this problem. We tried to use a statistical gaussian-interpolation technique to tackle the problem. Gaussian Process is the non-parametric method that tries to model the functional form of time series using a correlation matrix known as the kernel function. It not only models various crucial components of time series which are trend, seasonality, and stochasticity but also can be tuned to avoid overfitting and smoothen the time series using appropriate kernel functions which could be a problem with the naïve approaches such as interpolation which tries to fit a certain polynomial function. Since it is a statistical model every predicted/filled value there is an associated probability which gives the confidence of being an instance of the parent time series. We will also highlight how our method compares to the other commonly used method and instances where it can be deployed without risk. Often, some time-series datasets are rendered useless in our analysis/model due to the lack of availability of data at the preferred granularity. It is also sometimes not feasible to capture the data at a given frequency or only some sample data at the frequency it is available. Our method deals to tackle exactly those problems by providing an accurate estimate for the underlying data.
    Audi 3

  • Only 40% (chemo) - 54% (targeted therapies) cancer patients benefit from current treatments. Accurate prediction of patients' responses to therapies will significantly improve treatment outcomes. We have developed brAIn, an AI Platform trained on millions of clinical and molecular datapoints to accurately estimate responses of individual patients to >30 chemo and targeted cancer therapies. brAIn is validated on retrospective data from >30,000 patients (>100 clinical trials) and achieves >80% median sensitivity and specificity against multiple clinical endpoints and provides molecular-matched cancer drug recommendations with supporting evidence (evidence-based medicine). We are setting up brAIn as a SaaS framework to support oncologists by enabling timely diagnostics and treatment recommendations for cancer patients.
    Audi 2

  • Patients’ increasing digital participation provides an opportunity to pursue patient centric research and drug development by understanding their needs. Drug development can benefit from including patient and caregiver social media listening (SML) as part of patient-focused drug development (PFDD) strategy. This work introduces an Artificial Intelligence (AI)-enabled SML concept to identify and understand patient and caregiver experience to inform characterization of unmet medical need, target product profiles, value demonstration strategies, evidence generation strategies, definition of patient Centered outcome (PCO) endpoints, and selection of PROs in trial design.
    Audi 1

  • From Una in Himachal Pradesh to Tuticorin in Tamil Nadu and Mundra in Gujarat to Guwahati in Assam, the 15,113 km based massive network of pipelines in IndianOil is engaged in the safe and sustainable transport of crude oil, petroleum products and natural gas. Apart from the natural vulnerabilities, including corrosion and leakages, the pipelines are also subject to various intrusions by miscreants in an attempt to steal products. Each of these vandalism attempts could account for damages in crores. Apart from economic losses, any pilferage-led spill endangers human lives and the ecological viability of the surroundings, along with threatening the energy security of the nation. Catering to the prevention needs of pipeline integrity against pilferages, a machine learning-based model, the Pipeline Vulnerability Management System (PVMS), has been built. PVMS is an ensemble-based classification model that identifies certain stretches of pipelines to be more vulnerable (Red Zones) to pilferage attacks than the rest. The limited security personnel can be optimally deployed in the Red Zones to prevent pilferage events. The solution assures extensibility to various other industrial applications (Heatmapping Conveyer belt maintenance, logistic supply chains in FMCG), and further research can help enhance assessment at a more granular level.
    Audi 3

  • In an Omni channel marketing setup, Customer experience various ads via multiple channels before the final conversion. In a last-touch model one may miss the insights regarding the contribution of marketing channels that appeared in the customer journey prior to the conversion. A best-in-class Marketing Channel Attribution Model was developed using Markov’s chain to arrive at contribution of each channel and to identify cost optimum channel to be used for awareness and conversions.
    Audi 2

  • Large scale Machine Learning solutions tend to fit the “head” - needs of the more engaged audience. The performance can be poor for the under-represented populations whose votes get lost among the vast data. In this talk, we will discuss the initiatives underway in the internet product industry to democratise machine learning models for creating value for the whole population.
    Audi 1

  • Today’s data centres have thousands of servers that belong to multiple vendors, and they vary vividly in terms of technology, model, generation and numerous peripherals like firmware, software, and other hardware components. It is a challenge for the data centre administrators to track activities being executed across these servers and the resource consumption at various stages of deployment, delivery, installation, and upgrades. This leads to a subsequent ripple where the data centre administrators find it tough to foresee GPU availability and/or consumption in their data centres. This poses a risk when they need resources to be available for critical activity in the future, but because of an overburdened data centre, the resources are unavailable, and the critical tasks fail. On the other hand, if the data centre always remains underutilized, the unused servers in the data centre become a liability and bring the ROI down. This not only leads to unmitigable risks but also has a monetary impact on the profitability of the organization. This also leads to poor resource planning because of the lack of visibility of future needs. This paper focuses on designing a forecasting solution for data centre admins. The solution includes performing data explorations and developing enough understanding of the various metrics of the servers in the data centre like CPU, IO, Power etc., and developing univariate/multivariate forecasting solutions using the time series data sets of these metrics. This paper proposes to explore TDM as well as GAM models for predicting behaviors of various factors of data centres and CSPs. This solution would assist the data centre admins to have clearer visibility of the future load/availability of the GPU in their data centres. By having so, they would be in a better state to understand the utilization metrics for their data centres and better financial planning for future investments based on the clear visibility of under/over-utilized resources. This will also lead to better data centre resource optimization and rationalization.
    Audi 3

  • Semiconductor and data storage are highly complicated and capital-intensive manufacturing. It often requires Industry 4.0 solutions to target key business metrics such as reducing manufacturing cost, improving capital efficiency, reduce time-to-market to develop new products. The solutions span around AI/ML, Digital Twin, and Operations Research which originates as first-of-a-kind with an intent to take it scale. As India commits to catalyzing the semiconductor ecosystem in the country and many companies invested, join me in learning the challenges or gaps in skills and techniques on complex use cases in Advanced Analytics to propel the vision forward.
    Audi 2

  • Businesses increasingly focus on brand loyalty and customer engagement as they move towards more customer-centric strategies. Besides end consumers, sellers also form a key customer base for eCommerce marketplaces. Bad customer experience can lead to negative brand perception and drive down NPS (Net Promoter Score). This study was to help an eCommerce Marketplace scientifically identify key experience areas that were leading to a drastic decline in Seller NPS scores across regions over eight months. This would help businesses focus on the most important levers, prioritize budgets and improve NPS. In this study, we modelled customer survey data on brand perceptions and experiences, with historical transactional data and company policy information, to determine drivers of NPS. It was important to understand the inter-relations among various aspects of customer data, such as brand perceptions, issue resolution, customer service and macro factors, to provide holistic and actionable recommendations. We used a path model approach with a Bayesian network to determine causality between the key factors driving NPS and the interrelation among each factor. Structural Equation Modeling followed this to quantify the relative impact of each factor on NPS. The study found that sudden policy changes without effective communication, poor customer service and lack of protection for sellers from false buyer claims led to a poor selling experience. Sellers perceived the site as unfair to them and unsafe to sell. These insights were actioned upon by the company in an organization-wide cross-team initiative which helped address the seller’s dissatisfaction and eventually brought NPS back on an upward path.
    Audi 3

  • Today AGI is changing the world and Human expectations from the machine. Increasing variety and complexity of human-machine networks is failing non-deterministic behaviours of both the end. We may need to focus on two major aspects: Trust building in the new agency method Perceived risk and probable mitigation
    Audi 1

  • A talk on how to deal with data imbalance in machine learning/deep learning classification problems. We will discuss classical machine learning techniques of Oversampling, Undersamping, hybrid techniques, and cost-sensitive learning, as well as some advanced deep learning techniques like Dynamic sampling using PyTorch. If time permits, we will discuss how data imbalance affects the calibration of models.
    Audi 2

  • Executives across all industries have recognized the impact that Artificial Intelligence and Machine Learning can have on optimizing every part of their businesses. So why are so many companies struggling to capitalize on the power of these capabilities? The answer often lies not in the effectiveness of the algorithm but in the disconnect between the teams that manage customer experience and the data science/engineering functions in large organizations. In this talk, we will explore a few examples of successful and challenged deployments of AI and offer a perspective on how to deliver game-changing AI by starting with the customer, not the math.
    Audi 1

  • Day 2

    Jan 20, 2023

  • Advancements in technology, data and analytics are enabling predictive modeling and forecasting to help improve health outcomes. Similar to how weather forecasting proves to be very useful to prepare for uncongenial weather conditions, disease forecasting can be used to mobilize, allocate and manage resources as well as inform the general population of the impending onset of a seasonal disease, such as the flu, in a timely manner. In the coming years, disease forecasting is positioned to become more accurate and meaningful, with richer and better sources of data from wearable devices, improved ambient computing, more readily available gene data etc. Forecasting methodologies and data analytics models are also advancing, with ongoing innovations around recurrent neural networks, compared to traditional statistical techniques. All of this is powered by the awareness of populations for more accurate information on seasonal and chronic diseases.
    Audi 1

  • In this technical talk, we introduce a method for improving topic modelling by incorporating zero-shot classification with generative AI. Our approach considers the top words of a topic and the contextual meanings of the documents themselves to identify relevant documents within the topic. The use of this technique can have significant business impact, including the ability to more thoroughly analyse a topic and gain actionable insights relevant to each top-word.
    Audi 2

  • Most engineering product improvements are driven based on feedback from users and engineers. B2C products, such as the ones used to target customers or send personalized communications or manage order requests, track event-level actions and failures to improve product performance. However, the volume of failure logs (often in the order of a billion) and their unstructured nature (machine logs with minimal friendliness for human understanding) often hinder the detection of underlying themes from event failures. This paper discusses a unique and highly efficient approach to tune and leverage a language model for embedding generation. Using a weighted clustering technique, the embeddings are subsequently used to group failures into auto-detectable themes. The paper also proposes distinctive methods to manage embeddings that help improve the algorithm's performance, while retaining its focus on efficiency and computation time. Our experiments show that the proposed technique provides similar performance to the latest language models while taking less than one-tenth of the overall computation time.
    Audi 3

  • The talk will elaborate on building an end-to-end AI deployable pipeline for various industries (Health, Smart Spaces, Autonomous Driving, Industry 4.0, Conversational AI, Retail, Telecommunication, Energy) that can be brought to live-production in optimal cost and efficient timeframe for businesses.
    Audi 1

  • Traditionally, retailers identified similar store clusters based on demographic and physical attributes. They are ingested into tools like Planograms to determine the right assortment, develop aesthetic merchandising strategies or run mass-scale promotional offers across similar stores. The approach has an inherent focus extrinsic to consumer needs and more on parameters like location and physical attributes of the store like region, store size, shelf space etc. It completely neglects product demand within a store. This paper aims to develop intelligent store clusters prioritizing consumer demand for categories. Firstly, it provides a comparative view of coverage versus contribution perspective on how best to create category- wise store performance data. Secondly, it employs the ensemble clustering algorithm on this dataset that outputs high-accuracy partitions of performance-driven clusters. Using a normalized similarity matrix as a consensus function to combine results from multiple clustering models yields a more robust view of frequently grouped stores. It demonstrates ensembles’ increasing usability and supremacy within the unsupervised space too. Finally, it provides insights into indivisibility observation, where the groups could not be further broken down even after forcibly increasing the number of clusters. It highlights the potential of this approach to set up regional boundaries within a nation based solely on consumption patterns independent of social, political, and economic causes.
    Audi 3

  • 1. Across industries, organisations realise the strategic importance of data to drive growth. There is a sense of urgency to invest in systems that can generate timely and relevant insights for gaining an edge in the marketplace 2. In today’s digital world, data generation and its use cases have grown exponentially. With different forms of data, the applications for machine learning are also ever-increasing in order to make data sensible and actionable 3. Despite challenges, there are multiple opportunists across industry verticals to further accelerate the use of machine learning and automation that, in turn enhances the RoI for data & analytics functions.
    Audi 2

  • Undeniably, customer experience has emerged to be one of the most critical business imperatives, especially in the post-pandemic period. As organizations rethink and reimagine their CX strategies, Customer Care Operations functions are struggling to embed a data-backed decisioning system in their processes, which not only requires harmonizing humongous amount of customer interaction data across channels and touchpoints, but also measuring the right metrics and leveraging the right analytics frameworks to unlock quick, accurate and actionable insights, every single day. In this talk, we will share an approach on developing a data led, analytics accelerator for Customer Care domain that stitches together reusable data engineering, analytics, and ML components to optimize Data-to-Insights-Action journey in customer contact centers.
    Audi 1

  • Machine learning has started playing a significant role in complex decision-making processes. Accelerating this, counterfactual explanations in model explainability gives a brief idea about "what could have been the possible output if the input to the model had been changed in a particular way". With this, one can understand user behaviour and for any recommendation or user conversion these explanations are very useful to get an estimate of which variable needs to be changed and by how much. Directive explanations for counterfactuals output provide the actions that are required for changing the input variable from state A to state B. These explanations are personalized in accordance with each user which makes this idea very unique. This paper includes a novel approach to getting directive explanations by using Markov Decision Processes and Reinforcement Learning.
    Audi 3

  • Financial Inclusion is a major lever for boosting the economy of developing countries. FinTech companies, leveraging AI and Big Data, have been playing a crucial but subtle role in promoting financial inclusion, by enabling Credit access for a large base of consumers, even when they do not have format credit or banking history. In this talk, we’ll look at how non-traditional data and innovative ML techniques help us assess credit quality and affordability of consumers.
    Audi 2

  • The modern Data Scientist or Machine Learning Engineer is increasingly being called upon to build data products end to end. In this talk we try to do just that. While it is obvious that building a complete product isn’t the job of a single individual, it is critical that data professionals are aware that it is indeed possible to build end to end data products with modern data technologies. We will examine open data sources such as Common Crawl internet corpus, a data asset that has found widespread use in the development of corpora for language models. We will follow the process of converting raw data into various structured formats and finally build a data product, all within the Python ecosystem. For a more engaging read, do check out the book https://buildadataproduct.netlify.app where I go into a lot more detail about building a data product.
    Audi 1

  • Federated learning helps one leverage AI/ML techniques while preserving the privacy of localized data. However, owing to its decentralized nature, Federated Learning faces several optimization issues. This paper identifies the problem of incoming network congestion concerning the Aggregator in a federated scenario and proposes a statistical significance test to address the problem. Further network optimization is done by implementing a requirement-based request-response communication architecture to reduce unnecessary training rounds. This research also targets the infamous bias problem introduced due to label bias at the clients in a cross-device FL setting. The proposed solution defines a novel biasing factor to tackle data bias in the Aggregated model per the principles of AI ethics without violating privacy norms.
    Audi 3

  • For large pharmaceutical organizations to provide a better quality of life to patients, they must interact with and provide services to patients with unmet needs. Hence, the agents at patient call centres of such companies need to be highly efficient. To measure the effectiveness of such agents, the inbound and outbound calls with patients must be evaluated closely. Doing this manually for every call, although effective, is time and cost-intensive. Hence, the supervisors can evaluate only a handful sample of calls manually. Could we imagine an AI solution that can assess the sentiments on the call and adherence to call guidelines throughout the duration of each call? The idea, though novel, has some major obstacles due to (a) challenges in correctly separating the sentences spoken by patients and agents; (b) PII (Personal Identifiable Information) and PHI (Protected Health Information) regulations on patient data handling in the pharma industry; (c) correctly identify patient‘s intent and emotion etc. In this paper, we shall demonstrate a deep learning-based Natural Language Processing (NLP) solution to identify speakers (agent, non-agent, etc.) from redacted conversations, derive patients’ sentiments through the call duration and agents’ response to each negative emotion, agents’ adherence to call guidelines including repetitions, long hold-times, etc. and finally develop a scoring mechanism to assess the agent’s effectiveness in handling the call.
    Audi 3

  • The success of Transformer models in NLP led people to explore its application in computer vision tasks like image classification, object detection, segmentation, generative modelling and video processing. The workshop will discuss various Transformer based architectures used in Computer Vision and in particular, cover end-to-end use cases for Image Classification using various Transformer architectures.
    Audi 2

  • Using Deep Reinforcement Learning for autonomous system control entails finding the right combination of parameters for achieving optimal quality. Finding global optimum is often challenged by several factors such as high dimensionality, non-linearity and noisy state and action spaces along with convergence issues. Conditional generative adversarial networks (C-GANs) are used to learn the underlying distribution of the solutions generated by the optimization algorithm, and then generating unseen, more optimized solutions to the original optimization problem using the generative model. These solutions are used to sample the episodes for training the agent in subsequent episodes thereby reducing the variance and enhancing the probability of more optimal solutions.
    Audi 1

  • Natural disasters devastate local communities and make search and rescue (SAR) slow for those directly affected. Especially in harsh conditions, the search and rescue of people can take several days, which may be life-threatening in some cases. Technology that allows faster analysis will improve search and rescue efforts both in search time and resource management. We developed a deep neural machine (DNM), a novel artificial intelligence technology that processes UAV imagery for detection of people and other objects of interest in harsh environments, specifically for aquatic and nautical search and rescue. The DMN architecture utilizes convolution, batch normalization, and MISH activation (CBM) blocks, cross-stage partial (CSP) network blocks for residual features, and spatial pyramid pooling (SPP) blocks for merging different levels of features. Augmented versions of various datasets with synthetic weather scenarios are used for effective training of the object detection system for enhanced performance in changing environments. The DNM based autonomous vision system provides robust detection of humans and other objects of interest in aquatic and nautical environments.
    Audi 1

  • In this paper, we present the method of building a scalable, real-time inference platform for large-scale time-series anomaly detection and root-cause analysis solutions built as a part of the AI For Operations (AIOps) tool. AIOps is a tool built to ease the manual and time-consuming activities of DevOps engineers involved in monitoring and troubleshooting production systems. Such a system has to be operated in real-time to detect anomalies in a plethora of time-series metrics and logs from the production systems in order to provide timely alerts and possible root causes for quick remediation and thus requires a low-latency operation. This system must be scalable for the vast amounts of data involved for ETL and ML inference jobs that the solution needs. In this work, we show how we engineered and scaled up the AI research POC to a solution that supports a massive search engine system, where we achieved a reduction in latency by 30x. We also evaluate different tools for inference, such as Apache Airflow, Serverless REST API and Spark engine and demonstrate our improvements achieved and our estimations of these different commonly used platforms for ML inference in terms of feasibility and cost for an AIOps solution.
    Audi 3

  • In the machine learning world, it is quite a common scenario to come across the issues of variability and heteroscedasticity. One of the ways to address this is to assign weights to the observations. This helps us to ensure that the outliers are given lesser weights. Thus, reducing their effect on the model. Also, in certain cases, it is required to add some linear constraints on the estimates to restrict the predictor to be greater or lesser than a certain value or within a given range of values. This is mostly driven by a deep understanding of real-world behavior and by certain business scenarios. Based on data analysis, it was found that market volume had an impact on the costs, which required us to weigh the observations proportional to the volume of the container movement over twelve months. Added to it, the algorithm of the model had to accurately reflect the higher cost higher for refrigerated containers over the containers for dry cargo, given that all other features remained the same. This is a clear case of the combination of a Linear regression model with weights and subjected to certain constraints. After an in-depth study of existing approaches, we, in this paper, explore the application of a weighted constrained linear regression model, which would address the issues of variability and adding constraints to the predictor based on our business scenario.
    Audi 3

  • Traditionally, each time series were treated in isolation and as a result, traditional forecasting has always looked at the history of a single time series along with fitting a forecasting function. But recently, because of the ease of collecting data, many companies have started collecting large amounts of time series from similar sources, or related time series. This opens up an opportunity for a new and exciting modelling paradigm – Global Forecasting Models. Instead of fitting a separate forecast function for each time series individually, we fit a single forecast function to all the related time series. This paradigm has proven itself in the industry as well as in forecasting competitions in the past few years. We will look at why Global Forecasting Models are revolutionizing time series forecasting and discuss ways to conceptualize and improve global forecasting models.
    Audi 1

  • Anomaly detection has received significant attention from the researcher/practitioner communities for a few years due to its ability to identify anomalous incidents in data which could help to mitigate future risks. Recent advancements in Big Data technologies have enabled to process of huge amounts of data. While working on Big Data (Sensor, Transactional, etc.), Traditional Anomaly detection techniques have certain limitations which can be resolved with state-of-the-art machine learning models. In this study, we give a thorough assessment of the literature on anomaly detection methods. There are both conventional and machine learning techniques which can be applied to several data types. We describe the most popular supervised machine learning algorithms, i.e., KNN, Neural Networks and unsupervised i.e., Isolation Forests and Support Vector Machines. We also present comprehensive case studies about the use of anomaly detection in the banking and manufacturing sectors. In banking, Anomaly detection can be effectively used to detect fraudulent activities such as Credit card fraud, anti-money laundering, etc. In Manufacturing, it can be used to detect abnormal behavior of machines using sensor-based data. Anomaly detection is extremely critical today to avoid potential future risks with early identification of anomalies as businesses deal with trillions of data points and millions of metrics.
    Audi 3

  • DevOps inspired the development of MLOps practices to automate and streamline the entire machine learning lifecycle, from development to production and maintenance. Now when businesses are heavily relying upon data-driven decisions, there is a need for automating the workflow to improve the quality of data, increase the efficiency of analytics, and reduce the time cycle of data analytics and machine learning. This workshop enlightens the MLOps practices and puts focus on the requirements of DataOps. Along with that, the DataOps methodology to create automated data workflow for organizations will be covered in detail in this workshop.
    Audi 2

  • Visualization of reinforcement learning (RL) environment and learning dynamics of an agent is a vital step for debugging and a better understanding of the learnt policy. For virtual game environments, it is possible to visualize agent’s performance by rendering game screens. But for environments with optimisation of real-world multidimensional spaces with continuous variables, such as optimisation of chemical process parameters, it is challenging and complex to observe agent’s behaviour with visualization. This field largely remains unexplored in the research community. In the current work, a reinforcement learning agent is developed for optimisation of the production pro- cess of rubber mix for the tyre industry. This paper presents an attempt to visualize an RL agent’s training and inference for high-dimensional state space problems with continuous state and action spaces. A number of techniques are presented here to assist in debugging and monitoring the convergence of an agent over the complex domain. We explore plots for studying simulation environment, RL training dynamics, analysing trained policy and performance evaluation of trained policy in a given environment. Techniques described here are developed for actor-critic algorithms but can easily be extended to any RL algorithm.
    Audi 3

  • AgTech is a burgeoning domain and holds a lot of potential for Artificial Intelligence and Machine Learning. Currently available data in AgTech is scarce, incomprehensive and inadequate. Training machine learning models on such data can lead to a subpar performance in real-world environments. This includes scenarios such as changing weather conditions, different soil types, growing crops, shadows, angle of sunlight, light intensity, etc. In this paper, we showcase different augmentation and image synthesis techniques to align the available data in hand with the actual deployment environments and, consequently, make the model generalize better in order to handle real world scenarios. We are achieving this with paired image-to-image translation GANs(Generative Adversarial Networks) like pix2pix to synthesize new images by either re-using existing masks or drawing new masks from scratch. We then simulate diverse projections to deal with camera angles and different directional data, including perspective transformation. Further, in order to augment our data with varied weather and lighting conditions, we are using Contrastive Unpaired Translation and state-of-the-art brightening models like ZeroDCE++. We then augment our data with different soil types and use SR-GAN(Super Resolution GAN) to make the images look more real. Additionally, we are leveraging the Stable Diffusion model to augment crop and weed classes found in the AgTech domain. We benchmark the efficacy of these efforts using qualitative as well as quantitative evaluation metrics.
    Audi 3

  • Query Autocompletion (QAC) is a common feature for input text-based applications where a user’s partially typed prefix input is completed. It has primarily been studied for applications involving search-based queries that are short sequences or phrases. We present a novel approach to QAC for a Question Answering system in an Augmented Analytics platform where queries are essentially business and analytical questions in natural language. Our approach involves a combination of semantic search and natural language generation via beam search for completing questions. To enable generative completion in natural language and handle unseen prefixes, we use a pretrained distilgpt2 model that is fine-tuned for question completion tasks. In addition, we describe a method to synthesize training data from limited available past queries for fine-tuning the model and generating quality results for completion. We evaluate the proposed method using two datasets from different business domains based on fluency, style of text and business sense of the application with perplexity and precision. Experimental results demonstrate that our framework is effective and generalizes well on unseen prefixes with acceptable latency using suitable optimizations.
    Audi 3

  • AI is one of the most dynamic technical fields. It becomes demanding for data science/AI professionals to keep up with the pace of the technology. But more than keeping up with learning demands, it becomes hard for data scientists to differentiate themselves as access to learning resources is highly democratized in the last 5 years. In today's market, the place where one studies or academic credentials are becoming less and less important. So how do data scientists prove their skills, showcase their quality and differentiate themselves, all while having a ton of fun? This session answers exactly that question. Join Abhijeet Katte, with: 1. Krishna Rastogi, Chief Technology Officer, MachineHack 2. Rahul Pednekar, Vice President - Advanced Analytics, SwissRe 3. Salil Gautam, Machine Learning Engineer, AirBnB
    Audi 2

  • Federated learning is a growing field in Data Science, still businesses are not much aware of it. Most of the businesses and organizations are concerned about privacy of the their data. Federated learning can be one such important technology here with which many businesses will be willing to integrate ML solutions without any data privacy concern. Along with privacy, FL is helpful in many other use cases. I’d like to explain integration of Federated Learning in various domains like manufacturing, pharma, healthcare, edge-devices etc. This talk will include the technical aspect of Federated learning and it will shed some light on business aspect too.
    Audi 1

  • A brain tumor can be deadly and may cause many severe issues and even death if not diagnosed and treated at the early stages. So, early detection of brain tumors is of prime importance. Meningioma, Glioma, and Pituitary tumors are the most common, and 73% (1) of brain tumors are diagnosed as one of these types. Brain MRI image is one of the crucial methods to diagnose a brain tumor. Manual reading of MRIs could be time-consuming, and interpretation may vary based on the reader’s expertise. So an AI-based automated computer-aided diagnosis (CAD) can help identify and classify various brain tumors. The extensive potentiality of AI has successfully solved innumerable complex problems in medicine. However, the lack of transparency persists due to the increased complexity of advanced Deep Learning models. The black-box nature of the models complicates the decisions of AI applications in clinical use. Thus to perceive the underlying high-stake decision-making processes and to make the models interpretable, explainable AI (XAI) can be used. This study has proposed a classification model based on EfficientNet- B5 with attention-based weighted Global Average Pooling (GAP) to classify three brain tumors. It also demonstrated the use of Explainable AI to visualize the affected region or area of interest identified by our black-box model. Transfer Learning has been used, and specific layers of pre-trained EfficientNet- B5 have been fine-tuned with an attention-based GAP layer. The proposed model achieved 93.73% validation accuracy in multi-class classification, which was 2% higher than the EfficientNet B5+GAP without the Attention layer. We have used Grad CAM to implement the visual interpretability of our classification model. Our model achieved a macro-F1 score of .94 on the validation dataset, indicating very high macro-Recall and macro-Precision. Still, if there’s any case of false positive or false negative, that can be identified by looking into the visual interpretation of the model predictions. The proposed method outperformed many similar studies and effectively explained the region of interest identified by our black-box model using Explainable AI.
    Audi 3


Grab your ticket for a unique experience of inspiration, meeting and networking for the machine learning industry

Book your tickets at the earliest. We have a hard stop at 1000 passes.

Note: Ticket Pricing to change at any time.

  • Regular Passes

    Available from 14th Oct to 24th Nov 2023
  • Access to all in-person tracks & workshops
  • Access the recorded sessions later
  • Access to in-person networking with attendees & speakers
  • Conference Lunch for 2 days
  • Group Discount available
  • 4000 + taxes
  • Late Passes

    Available from 25th Nov onwards
  • Access to all in-person tracks & workshops
  • Access the recorded sessions later
  • Access to in-person networking with attendees & speakers
  • Conference Lunch for 2 days
  • No Group Discount
  • 6000 + taxes