Nkanyiso22/Python-data-analysis-for-Insurance-payout-and-Electricity-usage

Part 1. Insurance plains associated with weather patterns in Nigeria with insurance claims Part 1: Implementing a well-defined prediction task. The task is to develop a machine learning model to predict whether a building in Nigeria will have an insurance claim in a particular period. The dataset is available in the form of an open competition hosted by Zindi (zindi.africa). There is no prize money associated with the competition; the prize is in the form of knowledge gained through taking part in the challenge. To access the data, you are required to join Zindi (click on “Join Zindi" at zindi.af rica). After signing in, you should enrol for the “Insurance Prediction Challenge" competition (zindi.af rica/competitions/insurance-prediction-challenge/data) as an individual so that you can download the data and agree to the conditions of accessing the data. The task for this part of the project is to perform exploratory data analysis on a complex realworld dataset. The aim is to give you the experience of trying to extract meaningful insights from data without any pre-defined data analysis question or task. Two licensed datasets are made available to you from a study of domestic electrical consumption over two decades in South Africa. The project and data is described in a document provided (see DELSD_for_SA.pdf). The two datasets provided are: • DELSKV: socio-demographic survey data of households for which metered electrical consumption data is also available. The questionnaire of the survey is included as a pdf with the data file so that you can see how the data was collected. You can access additional metadata via the repository: datafirst.uct.ac.za x.php/catalog/758/study-description. • DELMH: Hourly data of domestic electrical load metering, 1994 – 2014. This data is of the same households in the DELSKV dataset, so can be linked for joint analysis of the socio-demographic survey data. You can access additional metadata via the repository: datafirst.uct.ac.za.