#keras

I Tried to Prevent Overfitting. My Model Got Worse. Here’s What I Learned.

I Tried to Prevent Overfitting. My Model Got Worse. Here’s What I Learned. A student’s honest walk through building a digit-recognition neural network with Keras — including the part where the …

CNNs, RNNs, and Transformers Explained: A Mental Model for Key Deep Learning Concepts

Bài viết giới thiệu khái niệm cơ bản về neural network, deep learning và so sánh ba kiến trúc quan trọng: CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks) và Transformers thông qua mô hình tư duy trực quan.

Lập trình viên nên đọc bài này để hiểu rõ cơ chế hoạt động của các mô hình AI như CNN, RNN, và Transformer, giúp họ xây dựng và tối ưu hóa các giải pháp deep learning hiệu quả hơn trong thực tế ứng dụng.

Python and LSTM neural network models and CSI 300 and CSI 500 stock index forecasts | includes AI agents, code, and data

A walkthrough of building an LSTM neural network model in Python (using Keras) to forecast the CSI 300 and CSI 500 Chinese stock indices. Covers 16 years of daily trading data (2008–2024), data preprocessing with Z-Score normalization, a single-hidden-layer LSTM architecture with 50 neurons, Adam optimizer, Dropout regularization, and RMSE/MAE evaluation. The model achieves ~2.11% MAPE on CSI 300 and ~2.41% on CSI 500, and generates 10-day forward price forecasts. The article also discusses model limitations (missing macro/sentiment features, black swan sensitivity, manual hyperparameter tuning) and improvement directions including attention mechanisms, CNN-LSTM hybrids, and ensemble learning.