Missing Data Prediction from Wireless Sensor Networks

Abstract

The aim of this project Missing Data prediction is to predict the missing values of well-known organization called British Petroleum (BP) and to eradicate the noisy data from their datasets. This will aid to find out the future left over reservoirs and future oil prices . Various machine learning algorithms were implemented and the best one among them was chosen out which gives most accurate predicted values and lessens the noisy data. The algorithm implemented is molded into a module, which provides ease of use to the users and also predicts missing data in other data types such as Tensorflow Tensor, Pytorch Tensor, Numpy Array, MXNet Nd Array.A dataset of the same data types as of the input dataset by the user and with the missing values replaced by the predicted values of the prediction algorithm is the output of this module.