This project is a collection of recent research in areas such as new infrastructure and urban computing, including white papers, academic papers, AI lab and dataset etc.

The primary objective of this project is to build a Real-Time Taxi Demand Prediction Model for every district and zone of NYC.

Machine Learning Model for Order Demand Prediction based on historical Order data - Built for Swiggy Hackathon 2018

Integrated real-time data analytics for optimized public transport, innovative road monitoring using demand prediction, and conditioning tech for sustainability, real time pothole detection either by image or video, smart parking count system for efficiency using AI/ML.

this repository contains main project of Rahnema college machine learning bootcamp

A spare engine placement generator based on a Finite-Horizon Markov Decision Process

Online Product Forecasting Using Machine Learning is a machine learning-based system that predicts future product demand using historical sales data. It helps businesses improve inventory management, reduce stock shortages, optimize resources, and make data-driven decisions through accurate sales forecasting and trend analysis.

Dynamic Pricing is an application of data science that involves adjusting the prices of a product or service based on various factors in real time. It is used by companies to optimize revenue by setting flexible prices that respond to market demand, demographics, customer behaviour and competitor prices.

🅿️ Streamline parking management with our AI-driven platform featuring real-time tracking, user dashboards, and intelligent video analysis.

This repository contains an Excel-based dataset of original daily/monthly sales data intended for use in time series forecasting tasks. The dataset is suitable for training LSTM (Long Short-Term Memory) models and benchmarking forecasting performance.

Reproducible Python data pipeline for transforming Austin shared micromobility trip records into cleaned e-scooter dataset, hourly pickup/dropoff demand images, and geospatial activity mask for spatiotemporal deep learning.

Spatio-temporal traffic demand forecasting on a 15-min geohash grid. 42 leak-free features (target encodings, spatial-neighbour spillover, geo-clusters, cyclical time) + stacked ensemble of LightGBM, XGBoost, CatBoost, HGBR, ExtraTrees → 97.85 accuracy. Streamlit demo, geospatial viz, ablation study, error analysis, CI/CD, Docker, MLOps.

TimeSeries Analysis in R

Using machine learning methods to predict demand for bike sharing.

An AI-based inventory optimization system that leverages machine learning to predict demand, recommend menu items, and streamline stock management for restaurants and food service businesses.. — all deployed through a real-time Stream lit web app.

AI-powered Smart Metro Operations & Passenger Navigation Platform featuring QR ticketing, demand prediction, fraud detection, real-time analytics, and intelligent interchange guidance for Bengaluru Metro.

Retail sales forecasting project using ARIMA, Prophet, XGBoost, and LSTM — compares time series and machine learning models for demand prediction.

Dynamic pricing system for cement sales, using XGBoost and demand forecasting to optimize profit based on internal operations and external market conditions.

A research proposal on enhancing public transport operational efficiency using RWKV-based architecture with linear attention mechanism O(N), comparing against BERT Transformer Encoder.