---
tags:
- time-series
- regression
- svr
- stock-prediction
- scikit-learn
---
# SVR Model for AAPL Price Prediction

This repository hosts a trained **Support Vector Regression (SVR)** model and its necessary preprocessing components (MinMaxScaler) for predicting the closing price of **AAPL**.

## Model Details
- **Algorithm:** Support Vector Regression (SVR) with RBF Kernel
- **Features (Input Sequence Length):** 100 days
- **Target:** Single-step prediction (the price of day $T+1$)
- **Training Period:** 2020-11-24 to 2025-11-23

## Inference
To use this model, you must provide a sequence of the last **100** scaled closing prices.

1.  Load the `svr_model.joblib` and `minmax_scaler.joblib`.
2.  Scale your 100-day input sequence using the loaded `MinMaxScaler`.
3.  Run the prediction.
4.  Inverse transform the prediction using the loaded `MinMaxScaler` to get the final dollar value.