The landslides are a challenging problem in the Himalayan states such as Himachal Pradesh and Uttarakhand in India and as well as in the world. Machine learning models could be developed to predict the movement of landslides in advance. In our proposed study, we developed a univariate, multivariate, and ensemble multilayer perceptron (MLP) and trained on the data collected from ten different stations in the Mandi district in Himachal Pradesh, India. The primary goal of this paper is to develop a model to predict the movement value using time series data. Recorded data was divided into the 80:20 ratio to train and test the model’s performance. The root-mean-squared error (RMSE) was used to compare the model’s performance. In training, the multivariate MLP was the best model with 0.012 RMSE, and ensemble MLP was the second-best model with 0.025 RMSE. In testing, the ensemble MLP was the best with 0.012 RMSE, and the univariate MLP was the second-best model with 0.013 RMSE. The analysis of the results shows that ensemble MLP is a promising method that can be used for landslide prediction using movement data.