stl-decomposition

Time Series Forecasting Experiments A collection of hands-on experiments with time series data, featuring models like ARIMA, LSTM, and Prophet. From data preprocessing to forecasting, explore real-world applications like stock predictions and weather forecasting. Continuously updated with new techniques and models for better performance.

Time series regression modeling on a dataset of supermarket sales across years, with the Darts library in Python.

This notebook is designed to boost your day‑trading edge by providing an advanced, multi‑timeframe seasonality analysis

This repo applies STLplus to GRACE/GRACE-FO TWSA data from CSR (Center for Space Research) with a 0.25-degree spatial resolution and a one-month temporal resolution. STLplus is an improved decomposition approach of seasonal and trend decomposition using Loess (STL) by Cleveland et al. (2019), developed by Ryan Hafen.

Seasonal-Trend decomposition based on Loess + Machine Learning: Hybrid Forecasting for Monthly Univariate Time Series

Experimental MetaTrader 5 research dashboard for OHLC ingestion, PostgreSQL history, STL analysis, and review-gated trading workflows.

Telecommunication network traffic analysis and anomaly detection project for MCI Academy, leveraging time-series decomposition (STL) and Isolation Forest algorithms to predict network disruptions.

This project conducts signal decomposition on spatiotemporal data, such as hydrological data that varies spatially across grids over a specific period. The decomposition process is applied to terrestrial water storage anomaly (TWSA) data from the GRACE satellite mission.

Retail KPI anomaly detection + hierarchical root cause analysis (STL + robust MAD scoring) with a Streamlit dashboard, built on the M5 Walmart sales dataset.

A data analysis project using time-series decomposition (STL) and statistical tests in Python to investigate the correlation between weather and class attendance.

STL decomposition notebook with parameter analysis, diagnostics, seasonal strength metrics, forecasting, and alternative methods. Complete time series decomposition workflow with visualizations and statistical tests.

A Python-based portfolio optimization system using LSTM neural networks, ARIMA baseline models, and GARCH volatility weighting for Indian stock market analysis.

This repo contains the code used to collect, clean, analyze and model social media data to understand how communities grieve collectively in online social spaces. Full paper available.

STL, Isolation Forest, Z-Score detectors on AWS CUR data — severity-banded alerts + Streamlit dashboard. Cloud Computing term project.

A collection of practical R templates for time-series analysis, covering data preprocessing, exploratory analysis, diagnostics, forecasting, dimensionality reduction, classification, and segmentation. Built as clean, reusable reference implementations with documented workflows and standalone R scripts.

Multi-site industrial time-series forecasting and procurement optimization pipeline combining STL decomposition, Fourier multi-harmonics, and Bayesian-optimized gradient boosters.

Time series forecasting of TSX stock prices using ARIMA with STL seasonal decomposition in R

SARIMA forecasting on quarterly UK car sales (1977–2005) — Box-Jenkins methodology, STL decomposition, benchmark comparison, full PDF report