DeepLearning--LSTM-Stock-Predictor

Notebooks

Closing Prices Starter Notebook

FNG Starter Notebook

---

Data Preparation

• Concatenated Fear and Greed Index Values with Bitcoin Close prices
• Created function to slice data by window specified
• Split the dataset, 70% train, 30% test.
• Scaled feature and target variables

---

Building the LSTM RNN Model

• Specified three LSTM layers, each accompanied by a dropout layer
• Compiled and trained the model with 100 epochs and 1 batch size

---

Model Performance

• Evaluated the loss metric for each feature (Close price, Fear and Greed Index)
• Used the model to predict prices
• Plotted chart of real vs Predicted Price

---

Results

BTC Closing Model Prediction	BTC FNG Model Prediction