Machine Learning Projects for .NET Developers
Machine Learning Projects for .NET Developers shows you how to build smarter .NET applications that learn from data, using simple algorithms and techniques that can be applied to a wide range of real-world problems. You’ll code each project in the familiar setting of Visual Studio, while the machine...
| Autor principal: | |
|---|---|
| Formato: | Libro electrónico |
| Idioma: | Inglés |
| Publicado: |
Berkeley, CA :
Apress
2015.
|
| Edición: | 1st ed. 2015. |
| Colección: | Expert's voice in .NET.
|
| Materias: | |
| Ver en Biblioteca Universitat Ramon Llull: | https://discovery.url.edu/permalink/34CSUC_URL/1im36ta/alma991009629666606719 |
Tabla de Contenidos:
- Contents at a Glance; Contents; About the Author; About the Technical Reviewer; Acknowledgments; Introduction; Chapter 1: 256 Shades of Gray; What Is Machine Learning?; A Classic Machine Learning Problem: Classifying Images; Our Challenge: Build a Digit Recognizer; Distance Functions in Machine Learning; Start with Something Simple; Our First Model, C# Version; Dataset Organization; Reading the Data; Computing Distance between Images; Writing a Classifier; So, How Do We Know It Works?; Cross-validation; Evaluating the Quality of Our Model; Improving Your Model
- Introducing F# for Machine Learning Live Scripting and Data Exploration with F# Interactive; Creating our First F# Script; Dissecting Our First F# Script; Creating Pipelines of Functions; Manipulating Data with Tuples and Pattern Matching; Training and Evaluating a Classifier Function; Improving Our Model; Experimenting with Another Definition of Distance; Factoring Out the Distance Function; So, What Have We Learned?; What to Look for in a Good Distance Function; Models Don't Have to Be Complicated; Why F#?; Going Further; Chapter 2: Spam or Ham?
- Our Challenge: Build a Spam-Detection Engine Getting to Know Our Dataset; Using Discriminated Unions to Model Labels; Reading Our Dataset; Deciding on a Single Word; Using Words as Clues; Putting a Number on How Certain We Are; Bayes' Theorem; Dealing with Rare Words; Combining Multiple Words; Breaking Text into Tokens; Naïvely Combining Scores; Simplified Document Score; Implementing the Classifier; Extracting Code into Modules; Scoring and Classifying a Document; Introducing Sets and Sequences; Learning from a Corpus of Documents; Training Our First Classifier
- Implementing Our First Tokenizer Validating Our Design Interactively; Establishing a Baseline with Cross-validation; Improving Our Classifier; Using Every Single Word; Does Capitalization Matter?; Less Is more; Choosing Our Words Carefully; Creating New Features; Dealing with Numeric Values; Understanding Errors; So What Have We Learned?; Chapter 3: The Joy of Type Providers; Exploring StackOverflow data; The StackExchange API; Using the JSON Type Provider; Building a Minimal DSL to Query Questions; All the Data in the World; The World Bank Type Provider; The R Type Provider
- Analyzing Data Together with R Data Frames Deedle, a .NET Data Frame; Data of the World, Unite!; So, What Have We Learned?; Going Further; Chapter 4: Of Bikes and Men; Getting to Know the Data; What's in the Dataset?; Inspecting the Data with FSharp.Charting; Spotting Trends with Moving Averages; Fitting a Model to the Data; Defining a Basic Straight-Line Model; Finding the Lowest-Cost Model; Finding the Minimum of a Function with Gradient Descent; Using Gradient Descent to Fit a Curve; A More General Model Formulation; Implementing Gradient Descent
- Stochastic Gradient Descent
