davidwabuko/Tnzanian_Waterwells_Project

Classification Of Waterwells in Tanzania. ## Tanzanian Water Wells Project. ( ) # Brief Description The project entails: Business problem, Data, Methodology,Outcomes and Recommendations. # THE BUSINESS PROBLEM. - Tanzania has been facing problems with Water supply due to failure of Water pumps within their wells, I will create a model that is will predict water wells that require repair. Recommendations derived are presented to the Tanzanian Government and NGO's responsible for immediate actions and solutions. # METHODOLOGY. - Data cleaning and preprocessing: Handling missing values, outliers, and categorical variables. - Feature engineering: Extracting new features from existing data. - Model selection: Choosing an appropriate machine learning model for predicting water well conditions. - Model evaluation: Evaluating the performance of the chosen model. - Model deployment: Deploying the model to predict water well conditions. ## MODELLING # Decision Tree Model - The Descesion Tree Model I have come up with has shown A level of accuracy higher than that of the other models in that the rate of True positive determined was Higher than compared to my Logistic Regression model. # Logistic Regression Model - The Logistic Regression Model I came up with had a problem converging thre coeeeficients and therefore would have come up with a highly Overfitting or Underfitting results. ## CONCLUSIONS AND RECOMMEDNTATIONS 1. I recomend my Decesion Tree Model For its preciseness in locating Waterwells Location and Functionality Status of the Pumps within the wells 2. Through knowledge in this area I reccommend the use of Non-submissed Pumps because they break down faster, a factor caused by a drop in level of Water in the well hence they tend to suck up and pump mud causing clogging issues. 3. The most Probable cause of failing Pumps is as a result of Poor maintainance, I would recommend reshuffling the maintainance groups or deploying new maintainance groups …