Analisis Peramalan Harga Saham Menggunakan Temporal Convolutional Network: Studi Kasus PT Lippo General Insurance Tbk

Abstrak

The stock market has an important role in the Indonesian economy, but share price fluctuations are often difficult to predict accurately. The machine learning algorithm for forecasting stock price movement trends uses a Temporal Convolutional Network (TCN). This method uses a more comprehensive dataset and advanced analysis techniques to capture non-linear and dynamic patterns in stock price data. This research aims to predict the share price of PT Lippo General Insurance Tbk using Temporal Convolutional Network (TCN) to provide a more accurate and reliable forecasting model. The research method uses a quantitative approach with daily historical stock data from 2011 to 2023 which is processed through several stages, including data collection, pre-processing, model development, and performance evaluation.  The results of the study show that the stock price forecasting of PT Lippo General Insurance Tbk using the Temporal Convolutional Network (TCN) method produces values that are relatively close to the actual ones with MSE, RMSE, MAE, and MAPE indicators, respectively, being 11,076.8214; 105.2464; 63.5915; and 2.2369\%. This indicates that the TCN model is able to capture complex temporal patterns in the stock price data of PT Lippo General Insurance Tbk. The forecasting results that have been projected for the next 60 days, that the stock price of PT Lippo General Insurance for the next 60 days will tend to decrease from August 31 to November 23.