This research aims to predict the sedimentation rate in Padma River and to illustrate the rate of accuracy for water and land classification based on various machine learning algorithms. Increased sedimentation can cause utterly problems for the organisms living there. Moreover, the water flow could be changed; the depth of the water can be lessened because of river sediment deposition, making boating and recreational use extra tough. This study comprises sedimentation for the last five years and how it will influence the surrounding environment, ecology, and socio-economy in the upcoming years, predict future areas that are subject to deposition, and show the accuracy of different machine learning algorithms for water and land classification. To conduct this research, we used a dataset consisting of images of Padma River (close to Naria in Shariatpur district of Bangladesh), collected from satellite, and run image processing techniques on it. We have obtained an accuracy of 97.5% for the image of 2018 and 99.17% for the image of 2019 using K-means. On the other hand, we obtained accuracy of 91.67% for images of 2018 and 95% for the image of 2019 using a self-organizing map (SOM). However, random forest (RF) shows the best results, having an accuracy of 97.5% for the image of 2018 and 100% for the image of 2019.