Data Science/Ml Research Intern
Bengaluru, Karnataka, India
• Engineered a custom Python module to automate processing of heart rate data,replacing manual Excel processes; reduced data processing time by 80 hours permonth, enhancing data accuracy by 25%• brought forth new extraction statistical features for providing input for predictionmodels which improved the overall net model performance by 20%• Conducted extensive market research and brainstorming sessions to identifyinnovative methods for accurately predicting emotions from heart rate data andsports events; efforts led to a 15% improvement in predictive accuracy.• Developed improved and maintained a Sentient Analysis Machine Learning modelwhich was built using RNN and LSTM networks• Researched and brainstormed ideas for the structure of the RNN network andobtained a model accuracy of 83% on the accumulated user dataset• Did regular optimizations and modifications when new data and featureswere brought forth, reducing computational power by 5 %