mahin-arvind Goto Github PK

appliance-energy-prediction

bot-prediction-kaggle-challenge

faceemotionrecognition

Face emotion recognition technology detects emotions and mood patterns invoked in human faces. This technology is used as a sentiment analysis tool to identify the six universal expressions, namely, happiness, sadness, anger, surprise, fear and disgust. Identifying facial expressions has a wide range of applications in human social interaction detection for industries like digital learning and market research. In this project, the Face Emotion Recognition 2013 Dataset was used to train five different types of architectures built using convolutional layers. Three custom convolutional networks were designed using traditional convolutional layers, VGG blocks and a simple inception block. Apart from this, transfer learning was administered on VGG-16 and ResNet-50 architectures using ImageNet weights. A sum total of five models were trained and evaluated during the course of the project. The best-performing model was deployed as a Streamlit application to perform real-time emotion recognition.

gtd-eda

health-insurance-cross-sell-prediction

In this project, the goal is to use the Health Insurance Cross-Sell dataset to understand Vehicle Insurance Cross Sale Responses, apply machine learning techniques to identify Vehicle Insurance buyers among pre-existing policyholders and provide explanations from the best classifying model to understand factors affecting customer responses.

netflix-movies-and-tv-shows-clustering.

This project aims to cluster the video content available on Netflix based on the company’s site data. Apart from aiding in the development of an efficient recommendation system, clustering the video content would also provide information about the type of content the company is interested in listing on its site. Thus giving an insight to content creators and filmmakers on the type of video content in demand.

object-detection-app

Simple object detection app with streamlit

opencv

Open Source Computer Vision Library

rock-paper-scissors

sample-app-template

Use this as a template for creating new Sample Apps for AstraDB

seoul-bike-sharing-demand-prediction-capstone-project

In this project, the Seoul Bike Share Demand dataset was used to understand bike share use trends, apply machine learning techniques to predict the number of bikes rented at any given hour and provide reasonable explanations from the best predicting model to understand factors affecting bike share demands.

topic-modeling-on-news-articles