EARLY DETECTION OF EQUIPMENT FAILURE, SHELL India Markets Pvt Ltd
‣ Pre-processed and merged 3-years of time series Sensor Data of different Shell equipments. ‣ Applied Hidden Markov Model (HMM) based Sequential Clustering for diagnostics using hmmlearn library of python for pattern recognition and using that pattern to recognize the health state of the Asset by measuring the similarity of time series sensor signals.
2018-5 - 2018-10
ANALYSIS OF DRUNK DRIVING BEHAVIOR, IIT Kharagpur
‣ Used SensorLab Android application to collect accelerometer, gyroscope, GPS data of the driver and applied classification models. ‣ Used Support Vector Machine (SVM) classifier to train the model to distinguish between drunk driving and normal driving by integrating both driving performance and environmental data.
2018-1 - 2018-4
BEHAVIOUR BASED ANALYSIS OF MALWARE, IIT Kharagpur
‣ Extracted features using N-gram technique, applied classification models like Multi-Layer Perceptron(MLP), Random Forest & KNN. ‣ Employed Feature Reduction Techniques like Principal Component Analysis (PCA) and Random Forest's Feature Importance attribute.