Best practices for business contexts
by Joseph Babcock, Ashish Kumar
Python: Advanced Predictive Analytics
- Python: Advanced Predictive Analytics
- Table of Contents
- Python: Advanced Predictive Analytics
- Credits
- Preface
- 1. Module 1
- 1. Getting Started with Predictive Modelling
- 2. Data Cleaning
- Reading the data – variations and examples
- Various methods of importing data in Python
- Basics – summary, dimensions, and structure
- Handling missing values
- Creating dummy variables
- Visualizing a dataset by basic plotting
- Summary
- 3. Data Wrangling
- Subsetting a dataset
- Generating random numbers and their usage
- Grouping the data – aggregation, filtering, and transformation
- Random sampling – splitting a dataset in training and testing datasets
- Concatenating and appending data
- Merging/joining datasets
- Summary
- 4. Statistical Concepts for Predictive Modelling
- 5. Linear Regression with Python
- 6. Logistic Regression with Python
- Linear regression versus logistic regression
- Understanding the math behind logistic regression
- Implementing logistic regression with Python
- Model validation and evaluation
- Model validation
- Summary
- 7. Clustering with Python
- 8. Trees and Random Forests with Python
- 9. Best Practices for Predictive Modelling
- A. A List of Links
- 2. Module 2
- 1. From Data to Decisions – Getting Started with Analytic Applications
- 2. Exploratory Data Analysis and Visualization in Python
- 3. Finding Patterns in the Noise – Clustering and Unsupervised Learning
- 4. Connecting the Dots with Models – Regression Methods
- 5. Putting Data in its Place – Classification Methods and Analysis
- 6. Words and Pixels – Working with Unstructured Data
- 7. Learning from the Bottom Up – Deep Networks and Unsupervised Features
- Learning patterns with neural networks
- A network of one – the perceptron
- Combining perceptrons – a single-layer neural network
- Parameter fitting with back-propagation
- Discriminative versus generative models
- Vanishing gradients and explaining away
- Pretraining belief networks
- Using dropout to regularize networks
- Convolutional networks and rectified units
- Compressing Data with autoencoder networks
- Optimizing the learning rate
- The TensorFlow library and digit recognition
- Summary
- Learning patterns with neural networks
- 8. Sharing Models with Prediction Services
- 9. Reporting and Testing – Iterating on Analytic Systems
- Bibliography
- Index
This is the meatiest part of the report created for a predictive modeling project. Some users of the report will navigate directly to this section as they are primarily interested in the overall effect of the project. Thus, it is imperative to mention the highlights and most important findings of the project in this section. This is different from reporting the statistics, which is in a way the raw output of the predictive model. In this section, we will focus on the following:
- Findings and insights of the analyses
- Major problems identified
- Major results from the model
- The accuracy or efficiency of the model
- Action steps for the user to solve the business problem, and so on
If it is a customer segmentation problem, mention the names and characteristics of the segments identified along with the statistical summary for each segment. Recommend a plan to maximize sales and revenue (or whatever the business objective might be) for each of the segments.
If it is a regression/prediction/forecasting problem, mention the accuracy of the results along with a summary of the results. For example, the expected number of house sales in the coming year is around t (say 900K), according to the model. The accuracy of the model is a% (say 98.5%).
Don't write in paragraphs. Write in bullet points. Add relevant plots and graphs to summarize the results.
Tables are a great way to summarize a lot of information in a small space. Use a lot of them. Screenshots are also a great way to show results as they are quite widely used. Assumptions, if any, should be clearly stated.
-
No Comment