Anomaly Detection & Breakdown Prediction

Decision tree, Boosted trees, Statistics, Feature engineering techniques

Modeling machine breakdown using supervised learning over high dimension al time-series data.. Devices that capture multiple electro-mechanical features generate huge amounts of time-series data. We determined which set of features were significantly contributing to mechanical failure in different parts of the machine. Moving windows of time-series statistics were used to model conditions that represent the current state of the machine leading up to a future breakdown.

Targeted Extraction & Automated Understanding of Text

Parse tree, Graph traversal, Auto-encoder, generative sequence2sequence LSTM neural network

Identifying & extracting key concepts, questions in chat conversations reviews & recognizing values for domain attribute Processing text inputs can be challenging. NLP algorithms needed to work for multiple languages, cope with the use of slang and be able to process short-form text that has partial grammatical consistency. We used a combination of techniques involving parse tree traversals, a creation of overlapping n-gram sequences, language detection APIs and auto-encoder based pattern matching engine to extract concepts and questions from chat conversations or short reviews. These techniques can work for small data without the need for large trained models. Sequence2Sequence techniques were used to automatically convert a block of text into key-value pairs of attributes and their values. The final solution was an API-based service that was integrated into the main product.

Enabling Pragmatic Decisions with Price Estimation

Gradient boosted machines, gradient boosted tree

A Real Estate Investment Management startup wished to improve decision-making for investors using data-driven tools Factors that influence rental or sale value of homes depends not only on the structural parameters of the home but also on geographical parameters and chronological factors. Multiple models were created after feature analysis while accounting for partial data and in some cases very little data for a region. Boosting based algorithms for regression were employed and the core implementation was adapted for a case where a price range needs to be predicted instead of a point estimate. Thousands of models were created and deployed so that models suitable for a region can be used to make the predictions. The final solution was an end-to-end flow adapted for the product architecture.

Improving Accuracy with Deep Learning

Combination of deep auto-encoders, deep neural networks, support vector machines

A wireless network startup wanted to locate assets based on the strength of radio frequency accurately... We conceptualised a Solution keeping in mind all the constraints of infrastructure deployment. The new algorithm was developed using a combination of neural networks and statistical techniques, and was groomed to work harmoniously with the existing algorithm. This improved accuracy of the services for marquee users from domains like health-care, hospitality, retail etc.

Multimedia Trend Prediction with Segmentation

Clustering & instance-based algorithms

A digital media company wished for a solution that can predict the type of highly-popular content in various user demographics... We developed a solution to predict the percentage of possible views on multimedia platforms like YouTube, Facebook, Instagram, etc. coming from each of the demographic buckets. This helped the customer in identifying their target audience as per content genre and tracking popular content trending across geographies.