Prof. Dr. Alper FIDAN2025-10-142025-10-14https://dspace.academy.edu.ly/handle/123456789/1803The effectiveness of all machine learning models used in this research to predict missing logs has been proved. However, there is a slight difference in the prediction accuracy between these models. The prediction ability of the random forest algorithm leads in this research with a prediction accuracy of up to 99%. Then it is followed by the decision tree and the k-nearest neighbor algorithms. In respect of deep learning models, the ability of the model using the RELU activation function is superior to its counterparts by prediction accuracy of 99%, followed by the model using the GELU activation function, and the model using the SWISH.Artificial intelligence techniques, including machine learning and deep learning, have become widely used in many fields. One of the most important of these fields is the oil and gas industry. The aim of this research study is to produce predicted petrophysical data that compensate for missing ones. In this research, machine learning algorithms applied, namely the decision tree (DT), K-nearest neighbor (KNN), and random forest (RF) algorithms. Distinct deep learning models have also been designed using dense layers with different activation functions, namely Rectified Linear Unit (RELU), Gaussian Error Linear Units (GELU), and (SWISH) activation functions.APPLICATION OF MACHINE LEARNING AND DEEPAPPLICATION OF MACHINE LEARNING AND DEEP LEARNING TECHNIQUES TO PREDICT MISSING PETROPHYSICAL DATA