Mrunal Jadhav Email & Phone Number

Toronto, Ontario, Canada

Toronto, Ontario, Canada

• Analyzed behavioural data of platform users seeking patterns between price movements in an asset and reaction of traders towards it.
• Paired existing volume-based strategy signals with momentum alpha factor to reduce factor turnover and lowering potential trading costs. Achieved a 8% reduction in historical drawdowns and a 23% decrease in the number.
• Leveraged signal processing techniques including empirical mode decomposition to address non-stationarity in price data. Decomposed BTCUSD closing prices to identify four dominant intrinsic mode functions (IMFs) and.
• Trained an LSTM network on alpha factors and lagged returns of BTCUSD computed on identified IMF frequencies as features to forecast return direction on different lookahead periods.

Alberta, Canada

• Procedural content generation via machine learning techniques are applied to a handful of games due to limited availability of annotated datasets and inconsistent level representation practices across different games.
• Presented a representation learning approach for game level design thus reducing reliance on manual translations and domain expertise.
• Implemented a deep autoencoder network to map pixel representation of game objects, their context and semantics to a continuous embedding vector called tile embeddings.
• Tested the ability of tile embeddings to serve as a representation for annotated and unannotated games by training an LSTM-based game level generator on their tile embedding representation.
• Assessed the playability of generated content using tile using an A* agent, reporting an average playability of 48%.
• Style of the generated content was also evaluated with respect to its leniency, linearity, enemy sparsity, density, and interestingness. The results implied that the generator covered the generative space well and.

Alberta, Canada

• Course: Introduction to File and Database ManagementKey Responsibilities:
• Conduct practical labs and provide guidance to 50+ students on course projects.
• Design course and lab material and grade exams and assignments, provide feedback to students on projects and conduct weekly office hours to review lecture material.What did I learn?
• Gained a stronger understanding of foundational concepts in computer data organization and information processing including entity-relationship model, relational model, relational query language: SQL, storage.
• Writing Python API for connecting Python code to a database and SQL queries to retrieve desired data from the database.
• Develop a conceptual data model and translate ideas into components like tables, relationships, queries, and views.

Toronto, Ontario, Canada

• Assisted in adding an explainability mechanism to existing deep neural classifier which evaluates a client’s financial history
• Developed an explainability mechanism that monitors the influence of each instance on the weights of a Deep Neural Network (DNN) during training. During inference, this mechanism allows us to pinpoint the most.
• Evaluated the model’s capacity to provide influential training examples by conducting a comparative study between our explanations and the nearest neighbours of an instance. The experimental classifier, trained on.

Edmonton, Alberta, Canada

• Created work-integrated learning experience by developing case studies on Linear Regression, Logistic Regression, and Support Vector Machines.

Master Of Science - Ms, Computer Science, 3.9/4.0

Courses: Procedural Content Generation via Machine Learning, Explainable AI, Probabilistic Graphical Models

Bachelor Of Engineering - Be, Computer Science, 3.91 / 4.00

Courses: Data Structures and Algorithms, Database Management, Operating Systems, Computer Networks, Discrete Mathematics, Machine.