Data Analytics Bootcamp | Data Analytics, Data Science, Data Visualization, Machine Learning
Minneapolis, Minnesota, United States
In a team broke down data into categories and created different plots to visualize the data. Weperformed a correlation check and logistic regression to form our final analysis. A SQL database was created including schema, an ERD, and the primary and foreign keys. Addedmultiple datasets to database. Formulated queries that pulled out different filtered data. Used SQL and Python’s SQLAlchemy to pull information from a database to make analyses, thenformatted the data to JSON to be used for an API. Used HTML/CSS/JavaScript to make website that was interactable with a dataset As part of a team, we found a dataset and transformed it so others could make an analysis. Wecreated an ERD and schema for the database then used Pandas to transform the dataset intothe multiple tables of the future database and loaded the data into the database usingSQLAlchemy. AS part of a team, we found a dataset and implemented that dataset into a website that wasinteractable. Used Leaflet to show how the metrics looked worldwide. We loaded data into aflask server to grab as a JSON and used a GeoJSON to find country lines. We used colorcoordination to show the varying ranges of metrics and summary table page that wasinteractable with different plots for what the data described. Used TensorFlow to make a neural network then complied and trained for test and train data. Itperformed a classification analysis to determine success. Then optimized it to try and getbetter than 75% accuracy. Used AWS review datasets fit them into a SQL database. Used ETL to fit data to schema andcreate new columns based on the given data. Datasets were then loaded into the SQL databaseand queries were made. Trained and tested different models on a dataset, the models were than optimized andvisualizations were made. A website was developed to show the results and visualizations thatcame from the models. A summary page was then made to go over how each model performedand results found.