- Google Hashcode – Ranked 458 all over world
- Part Student Exchange Program at Ural Federal University, Russia for 45 days
- Attended International Level Hackathon at Ural Federal University, Russia.
- CODESTORM Hackathon, REVA University – 2nd Runner up
Travel-Insurance claim Prediction
- Prediction of false insurance claims by customer.
- Used Kaggle Dataset for training.
- Used different technique for EDA (Exploratory Data Analysis) and trained the model using different classification algorithms like Random Forest Classifier, Decision Tree Classifier, Linear Support Vector Classification and Gradient Boosting with an accuracy of 75%
Handwriting Text Recognition as part of 3rd Year Mini Project
- Handwritten Text detection in images to convert it to digital text.
- Used IAM word dataset for Training.
- Used Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) for training and got an accuracy of 84 %.
- Presented paper IACIT 2020 Conference
(http://www.testmagzine.biz/index.php/testmagzine/article/view/8197/6189 ).
Facial Expression Recognition Using Keras
- The objective is to classify each face based on the emotion shown in the facial expression.
- Built and trained a convolutional neural network (CNN) in Keras from scratch with an accuracy of
90%.
Handwritten Word Recognition using Keras
- Trained on EMNIST by-class dataset (alphabets: a-z, A-Z and digits: 0-9).
- Word separation is done using Contour detection
- Trained the model using Convolutional Neural Network (CNN) with an accuracy of 87%.
Prediction of Solar Output
- Trained on Yekaterinburg Weather Data.
- Predicts the Solar output based on Solar irradiation and other weather features.
- Trained using Decision Tree Regressor with an accuracy of 93%
Employee Attrition Prediction
- Used HR dataset from Kaggle to train model to predict the employee churn.
- Trained the model with Decision Tree and Random Forest with an accuracy of 93% and 95% respectively.
