Design Thinker & Code Tinkerer | NIT Meghalaya Grad

Yash Bagoria  • Peerlist

-Android malware detection system using ML algorithms, including K-Nearest Neighbors (KNN) and Decision Trees
-Analyzed a dataset of application permissions (330 features) to accurately classify apps as benign or malicious
-Achieved accuracy of 0.89 with KNN classifier (k=3), showcasing strong predictive performance
-Utilized Decision Tree classifier to achieve an accuracy of 0.9375 reinforcing robust malware detection capabilities

A Permission-Based Malware Detection System for Securing And

-Developed a predictive model using the state-of-the-art LSTM architecture to assess health of XLPE cables
-Leveraged LSTM's ability to capture temporal patterns in sequential data for over 93% accuracy
-Utilized 5 different features like age, neutral corrosion and relevant factors for accurate predictions
-Demonstrated the model's efficacy in predicting cable health for 2500 power cables in testing

Asset Management in Power Grid using Machine Learning

-Implemented GAN architecture to generate images from CIFAR-10 dataset comprising 60,000 labeled images
-Developed a deep convolutional generator and discriminator to enable image synthesis and class differentiation

Class-Conditional Image Generation using Generative Adversar

-Developed an advanced Image Classification pipeline for Monument Recognition using state-of-the-art ResNet based models. 
-Leveraged Transfer Learning Techniques to achieve exceptional accuracy and efficiency in recognizing various monuments from the Monument Recognition dataset
-Employed Transfer Learning to fine-tune pre-trained ResNet models, allowing efficient feature extraction and leveraging the knowledge gained from large-scale datasets to excel in monument classification
-Resulted in a reliable image classification system for monument recognition