-
The Programme:
- Create mental models to clearly define business issues,
- Visualize and prepare data to improve efficacy of predictive models,
- Identify and implement a variety of predictive modeling techniques
- Apply predictive analytics and business intelligence to solve real-world business problems
-
Information about the Program -> Link course
-
Detailed information about the Program -> Nanodegree Program Syllabus
- In this part, I have been approached a structured framework for solving problems with advanced analytics and a range of appropriate analytical methodology based on the context of business problem.
- How to build, validate, and applyLinear regression models was also introduced detailedly to solve a business problem. Along with theory, I did the practical project for prediction the expected revenue of sales from a catalog launch.
- Specifically, a home-goods manufacturer wants to predict expected profits from a catalog launch. As a Business Analyst I must perform the appropriate analytical methodology to assess the relationship between sales and other factores including the response from customers to the new catalog. After that I will build a linear regression model to provide results, evaluate its significance and give a recommendation to management board. More detailted:
Project: Predict Sales for a Catalog Launch
- Data Wrangling is at the core of all data activity. In business analysis, we need to be familiar with the work of processing different data types, dirty data, and outliers and how to reformat data and join data from different sources together.
- To do so, the role of understanding data, the issues of data, data formatting as well as how to deal with them was introduced clearly.
- In the practical part, I do project Opening New Store. Specifically, a pet store chain is selecting the location for its next store with specific requirements from management board. I need to use data preparation techniques to build a robust analytic dataset. With the clean dataset, I build a predictive model to select the best location. More details:
Project: Opening New Store
- Classification models are a powerful tool for business analyst. With the understanding of binary and non-binary classification models, I can use them effectively to drive business insights.
- In this part, it is also introduced when and how to build logistic regression, decision tree, forest and boosted models, and how to use stepwise to automate predictor variables selection then score and compare models and interpret the results.
- In the practical part, I do the project Predict Loan Default Risk. Specifically, a bank recently received an influx of loan applications. I need to build and apply a classification model to provide a recommendation on which loan applicants the bank should lend to.
More details:
Project: Predict Loan Default Risk
- Making the best decisions that is an essential part of Business Analysis. Therefore, planning and executing the analysis of an AB test allow us to provide confident recommendations.
- In this part, there are not only the fundamentals of A/B Testing, but also how to design a A/B Test including selecting target and control units and variables, the duration of a test as well as which kind of Design Tests should be implemented.
- In the practical part, I do the project Test a Menu Launch. Specifically, a chain of coffee shops is considering launching a new menu. I need to design and analyze an A/B test and then write up a recommendation on whether the chain should introduce the new menu.
- More details:
Project: Test a Menu Launch
-
In this part, I need to use the most important techniques including Classification Models, Time Series Forecasting and Segmentation and Clustering to plan a significant expansion for a grocery store chain.
-
Using Clustering Models to split the stores into the optimal number of formats.
-
Classify the stores to each kind of formats.
-
Decide kind of ETS model or ARIMA Model, and choosing the best one to predict the sale.