Digit Recognizer using Keras (Deep Learning)
Classified different digits using a convolution neural network with the data size of 42000 images and also achieved 3% more accuracy as compared to the previously generated model and detected anomalies in the dataset and evaluated the digits that have a high chance of false classification.
Trip Fare Analysis (Predictive Analysis)
In this project created a statistical model to identify the top 5 clusters of pick up and drop locations on a base map and helped in identifying 10K new potential customers and also used a linear regression model to predict the trip fare for a particular trip.
Sales Data Analysis (Descriptive Analysis Project)
Designed an automated psychiatrist mobile NLP chatbot application for cognitive behavior therapy for more than 20K people inside VIT Vellore Campus. The working of the model is based upon the video and image analysis using a deep learning model to calculate the percentage of emotions on the human face.
Mental Health Emotion Classifier ( Deep Learning)
Designed an automated psychiatrist mobile NLP chatbot application for cognitive behavior therapy for more than 20K people inside VIT Vellore Campus. The working of the model is based upon the video and image analysis using a deep learning model to calculate the percentage of emotions on the human face.