Please enable JavaScript.

Coggle requires JavaScript to display documents.

Chapter 2 (Chapter 3 (Chapter 4 (Chapter 5 (Chapter 6 (Chapter 7 (Success,…

- - - - Chapter 6
        
        Chapter 7
        
        Success, Risk and Continuation
        
        When evaluating project objectives, a better understanding of success criteria will develop over time.
        
        Start small with clear questions and goals such as:
        
        Who will use the model?
        
        Is management on board with the project?
        
        Can the model drivers be visualized?
        
        How much value can the model produce?
        
        Support from Management is necessary, when decisions are made about model implementation into the information workflow. As well, if the model truly drives value, that value must be specified and communicated relative costs and benefits.
        
        1 more item...
        
        Define Prediction Target
        
        The prediction target is the behavior of a "thing" we need to know about the future.
        
        Identify Defining Characteristics: The first kind is classification, which predicts the category to which a new case belongs. Alternatively, we can use machine learning to conduct regression, that is, to predict the target numeric values.
        
        When working through a dataset for classification, we carefully examine the possible states in which a variable can exist and consider weather they can be simplified.
      - Decide on Unit of Analysis
        
        A unit (of analysis) is the what, who, where of our analysis. For each unit of analysis, there can be numerous outcomes that we might want to predict
        
        How to Determine Unit of Analysis: One way to identify the appropriate unit of analysis is to first think about what the prediction target is.
    - - Subject matter expertise constitutes deep experience in a specific domain such as accounting, marketing, supply chain, or medicine. It is also important for early identification of potential obstacles and/or opportunities.
        
        With subject matter expertise you can set realistic expectations for model performance. Form your own suggestions for data collection to know where relevant data resides, including external data. Bring an expert on board to insure the maximum effectiveness of the project.
        
        Without a Subject Matter Expert or SME the following options are: discussing the domain to be modeled with data science colleagues, interviewing SMEs, reading trade journals, reading internet sources on the subject, or even reading the definitions of available features found in data dictionaries around the web.
  - - - Outline your case in business language.
        
        Any proposed project should be evaluated against three criteria:
        
        Is the project statement presented in the language of Business?
        
        Does the project statement specify actions that should result from the project?
        
        How could solving this problem impact the bottom line?
- - - - Important areas of Machine Learning:
        
        Exploratory data analysis. The process of examining the descriptive statistics for all features as well as their relationship with the target.
        
        Feature engineering. The process of cleaning data, combining features, splitting features into multiple features, handling missing values, and dealing with text, to mention a few of potentially hundreds of steps.
        
        Algorithm selection and hyperparameter tuning. Keeping up with the “dizzying number” of available algorithms and their quadrillions of parameter combinations and figuring out which work best for the data at hand.
        
        Model diagnostics. Evaluation of top models, including the confusion matrix and different probability cutoffs.
        
        Auto Machine Learning improves the overall system.
        
        2.2 What Automated Machine Learning is Not
        
        AutoML is not automatic ML. Decisions by the analyst and a requisite skillset for evaluating the results of applying machine learning to any data set.
        
        2.3 Available Tools and Platforms
        
        Two-Types of tools and platforms: context-specific tools and general platforms.
        Context-specific tools are implemented within another system or for a specific purpose. AutoML focus is on customer interactions.
        General platforms: Those that are designed for general purpose machine learning, splits into two types: open-source and commercial. In the open-source category, tools tend to be developed by and for computer and data scientists and generally require knowledge of programming languages such as Python and R. Commercial platform is provided by a commercial vendor, presumably for a price.
        
        1 more item...