Case Studies and Projects in Machine Learning/EDA/DL

Keeping Inventory of spare in various service centre to the market demand is always a challenge as most service centres spends significant amount in spare parts inventory costs. In spite of this, availability of spare parts is been one of the problem areas.

This repository is a guide to using various time-series analysis models like ARIMA, SARIMA, and SARIMAX in forecasting the exchange rates of 5 currencies as compared to US Dollars

Improved the accuracy of Bitcoin stock price predictions on ARIMA model by reducing the seasonality factor. Achieved RMSE value of 68.99 after implementation of SARIMAX model to reduce seasonality.

Sales forecasting system prototype w/ dashboard using SARIMAX model

This project uses Random Forest and ARIMA models to predict daily gold prices with 97% accuracy. By cleaning and analyzing historical data (2016–2021), we created a model that provides actionable insights. Deployed with Streamlit, it offers real-time forecasting for investors and traders to stay ahead of the market.

This repository contains a collection of time series analysis and forecasting projects, featuring both classical statistical models and deep learning approaches.

In this section, we will estimate airline passengers using time series methods.

In this section, we will examine the Statistical Methods in time series analysis.

Predicting orders for Glovo

Software Implementation of a Tourist Flow Model

This project creates Sales Volume predictions based on historical analysis of product, sales, and purchase data using SARIMAX time series model.

Using time series modeling to forecast the top 3 zipcodes to invest in Washington state.

Development of An Automated Conflict Prediction System by State Space ARIMA Methods

Predict the apple stock market price for next 30 days. There are Open, High, Low and Close price has been given for each day starting from 2012 to 2019 for Apple stock.

AI-Driven Crime Forecasting Across Indian States — A pioneering machine learning project that harnesses time series modeling (SARIMAX, Ridge Regression) to uncover patterns and forecast crime trends using real-world multi-state temporal and socio-economic data.

This project explores multiple time series forecasting techniques, including Facebook Prophet, ARIMA, and SARIMAX, to predict airline passenger trends. It compares the effectiveness of each model in capturing seasonality and long-term patterns

Enhancing Decision Making and Prediction Optimization using the HybridFlow Forecast Model

No description provided.

Time Series Forecasting with SARIMAX and XGBoost : Chennai house price prediction

Top-Down Investment Strategy Optimization with Time Series Forecasting

Time Series Analysis of Airline Passenger Data, In this time series forecasting, taking data from kaggle site and applying ARIMA and SARIMAX model to evaluate seasional trends of passenger travelling via airlines.

This analysis seeks to forecast the daily number of car accidents in the county of Denver, Colorado using both traditional time series methods from the statsmodels library, as well as the Prophet forecasting framework as implemented in the prophet library.

Sarimax - Seasonal Auto Regressive Integrated Moving Average with eXogenous Regressors – Time Series Forcasting Model for forecasting amrodd future financial values

Hourly Energy Consumption

Shell_App is an integrated data analysis and modeling platform, featuring scripts for EDA, ensemble techniques, LightGBM, and SARIMAX modeling. Designed for comprehensive data processing, the repository also includes deployment capabilities via Streamlit for interactive model interactions.

Evaluating Naive Baselines and Simple Seasonal Models for Cloud Capacity Planning. A reproducible analytics engineering project using dbt, DuckDB, and Python to aggregate global VM demand and benchmark rolling-average heuristics against interpretable seasonal time series models using Snowflake’s Shaved Ice dataset for ICPE 2026.

This was a competition hosted by Zindi, about predicting the percent of areas burned by wildfires in Zimbabwe. This project emerged the winner.

An ads and marketing based company helping businesses elicit maximum clicks @ minimum cost.

Previsão da Demanda Mensal de Energia Elétrica com Série Multivariada (SARIMAX)