Ayalem/Economic-Intelligence-Platform-for-Moroccan-CPI

interactive economic intelligence platform that analyzes and forecasts the Consumer Price Index (CPI) in Morocco using statistical and deep learning models (ARIMA, Holt-Winters, Prophet, LSTM). The goal is to transform economic data into clear, actionable insights for decision-making and inflation monitoring. # Economic Intelligence Platform for Moroccan CPI** ## **Project Description** *this project is an interactive economic intelligence platform . It analyzes and forecasts the **Consumer Price Index (CPI)** in Morocco using statistical and deep learning models (ARIMA, Holt-Winters, Prophet, LSTM). The goal is to transform economic data into clear, actionable insights for **decision-making** and **inflation monitoring**. ### Dashboard Preview --- ## **Project Objectives** * Collect and structure CPI data according to **COICOP classification**. * Build a **short- and medium-term forecasting pipeline** using hybrid models. * Identify **similar consumption patterns** through clustering techniques. * Develop an **interactive web dashboard** to visualize trends, correlations, and predictions. --- ## **Methodology** 1. **Data Collection & Preparation** * Cleaning, interpolation,chaining and normalization. * Mapping COICOP codes to labels. 2. **Exploratory Data Analysis (EDA)** * Trend, seasonality, and category correlations. 3. **Predictive Modeling** * Model comparison: ARIMA, Holt-Winters, Prophet, LSTM. 4. **Clustering & Visualization** * Applying KMeans * Visualizing with heatmaps and correlation networks. 5. **Deployment & Interface** * Flask API exposing predictions. * React dashboard for interactive visualization. --- ## **Main Features** * CPI forecast (short-term). * Correlation analysis between COICOP categories. * Clustering * Interactive dashboards with dynamic visualizations. --- ## **Technologies Used** | Domain | Technologies | | ---------------- | ---------------------------------------------- | | Languages | Python, JavaScript | | Backend | Flask | | Frontend | React, TailwindCSS, Plotly.js | | Machine Learning | Scikit-learn, TensorFlow, Prophet, Statsmodels | …