This research work presents the aftereffects of the advancement of a predictive model for the expectation of stock pattern dependent on a historical data gathered from the Nigerian Stock Exchange. First, is to collect a historical dataset which would be used to develop a model with machine learning algorithm after the historical dataset is trained with the algorithm. It is observed that a thorough exploratory research and preprocessing of the dataset and exploring Python libraries can yield to the development of good a model to predict the trend of stock with minimal error. To test for the precision of the model the historical dataset was split into training set and test set. The model was then evaluated with the root mean square error (RMSE) to see how large or minimal the error is on the prediction, this was seen to be closer to zero which shows that the prediction of stock trend with respect to the dataset is sufficiently okay. With the flexibility of Python a model was developed with the linear regression algorithm and the coefficients and intercept of the model were retrieved.

Full text: IJEIR_2445_FINAL.pdf