Projects

PBPK Modelling and Simulation Studies

Mentor - Prof. Sunil Kumar Dubey

[Jan 2019 – May 2019]

  • Worked on a classification task using various classifiers for predicting diabetic and Alzheimer activity of a lead molecule.
  • Developed an Android application which recognizes Medicinal Plants and solves the issue of adulteration. Used computer vision and machine learning techniques for extracting features from a plant segment.

Drug Toxicity Prediction using Pharmaco-chemical properties of a drug

Mentor - Dr. Vaibhav A. Dixit

[Dec 2018 – Jan 2019]

  • Worked on a multi-label classification task using clustering techniques and various classifiers to predict the toxicity of a drug, given the physicochemical properties as input.
  • Developed a web-based tool for calculating Drug Toxicity Index, which tells us how much toxicity a drug will have, compared to other standard values.

Plant Disease Detection App using Convolutional Neural Networks (Alex Net)

Mentor - Prof. Sundaresan Raman

[Aug 2018 - Dec 2018]

  • Developed an Android application which extracts features and recognizes various patterns to detect plant diseases.
  • Used TensorFlow framework and conv. neural networks having Alex Net architecture for training the model to predict with 91% accuracy.
  • Extracted plant images using various segmentation techniques like contours (OpenCV), applying thresholds, color spaces and watershed algorithm.

Time Series Analysis for Physiologically Based Pharmacokinetic (PBPK) Data

Mentor - Prof. Manoj Kannan

[Jan 2018 - May 2018]

  • Implemented time series prediction model (using PyTorch framework) for pre-clinical univariate PBPK data generated in different compartments.
  • Worked on a mathematical model of physiological and pharmacokinetic processes for various compartments.

Microsoft Hackathon (Microsoft Azure Flight Prediction Challenge)

Project (APOGEE – Technical Festival), BITS Pilani

[Feb 2018 - Mar 2018]

  • Developed a machine learning model using azure platform so as to maximize the accuracy of performance of a flight being on time. The model was trained to get an accuracy of 99.3%.

Early Prediction of Neural Disorder

Project (APOGEE – Technical Festival), BITS

[Feb 2018 - Mar 2018]

  • Used regression ML algorithms and image processing techniques on T1 MRI brain images to improve the accuracy of diagnosis of neural disorders, remove the number of false negatives and for early prediction.

Hobby Projects

[July 2017 - Present]

  • My other hobby projects include image segmentation, Q-Learning on Atari Games, Web Crawler and few other projects centered on computer vision algorithm and machine learning.