The observations that are used in a method to build a model for estimating the response variable.

Training data

The training and test data are generated by a probability distribution over datasets called the data-generating process. We typically make an assumptions that the examples in each dataset are independent from each other, and that the training set and test set are identically distributed, drawn from the same probability distribution as each other. This assumption enables us to describe the data-generating process with a probability distribution over a single example. 

The same distribution is then used to generate every train example and every test example. We call that shared underlying distribution the data-generating distribution, denoted $p_{data}$. 

This probabilistic framework and the assumption enable us to mathematically study the relationship between training error and test error.

Data-Generating Process and Data-Generating Distribution (in Machine Learning)

This function randomly shuffles the dataset and splits off a certain percentage of the input samples for use as a training set, and then puts the remaining samples into a different variable for use as a test set. So in this example, we're using a 75-25% split of training versus test data. And that's a pretty standard relative split that's used. It's a good rule of thumb to use in deciding what proportion of training versus test might be helpful.
```
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
```
- Unless optional parameters test_size or train_size are passed in, the method will subset the training and test data randomly into 75% train:25% test
- random_state is an optional parameter used for reproducible output across multiple function calls

Train Test Split Function

- Data is divided 80% as training set, 10% as development set, and 10% as test set.
- Test set should be as large as possible, since a small test set may be accidentally unrepresentative.
- A good common practice is to find the smallest test set that gives us enough statistical power to measure a significant difference between models, since we also need the training set to include as much as data.

Common Practices of Train/Test Set Arrangements in NLP

A test dataset is a separate partition of data that is held out from the learning process and used strictly to evaluate a model's performance. Because performing well on the training data does not guarantee success on unseen data, the test set serves as a critical measure of the model's true ability to generalize.

Test Dataset

A validation dataset (or validation set) is a separate partition of data used to assess the quality of a machine learning model's solution. After training a model iteratively on the training data, its predictive performance is evaluated on this distinct validation dataset to measure how well it has learned the underlying patterns. This evaluation is critical for automatically tuning hyperparameters without relying on the training set, ensuring the model generalizes appropriately to unseen data.

Validation Dataset

- Training set: to train the model
- Validation set (development set): to tune the model parameters
- Test set: to test the performance of the model on unobserved data

University of Michigan - Ann Arbor

Claude

Supervised learning is a type of machine learning problem where the objective is to predict a designated unknown label based on known inputs. This approach requires a training dataset consisting of examples for which the ground truth labels are already known, allowing the model to learn the mapping from inputs to classifications.

Supervised Learning

Dive into Deep Learning

Which of the following are use-cases of supervised learning?

Methods of supervised statistical learning

- Classification Problems
- Regression Problems

Types of supervised learning problems

Use cases of supervised statistical learning

Which ones are true about Supervised statistical learning?

Which of the following are examples of supervised machine learning? Select all that apply.

Categories of supervised learning algorithms

A Basic Supervised Statistical Learning Workflow

Division of dataset in supervised statistical learning

Feature scaling greatly affects which of the following supervised machine learning methods? 

1. Supervised learning allows you to collect data or produce a data output from the previous experience
2. Helps you to optimize performance criteria using experience
3 Supervised machine learning helps you to solve various types of real-world computation problems.

Adavantages of supervised learning

1. Decision boundary might be overtrained if your training set which doesn't have examples that you want to have in a class
2. You need to select lots of good examples from each class while you are training the classifier.
3. Classifying big data can be a real challenge.
4.Training for supervised learning needs a lot of computation time.


Disadvantages of Supervised Learning

1. Before doing anything else, you need to decide what kind of data is to be used as a training set
2. You need to decide the structure of the learned function and learning algorithm.
3. Gathere corresponding outputs either from human experts or from measurements

Best practices for Supervised Learning


The Supervised Learning Workshop: A New, Interactive Approach to Understanding Supervised Learning Algorithms, 2nd Edition Paperback – February 28, 2020
https://www.amazon.com/Supervised-Learning-Workshop-Interactive-Understanding/dp/1800209045/ref=sr_1_1_sspa?dchild=1&keywords=Supervised+Learning&qid=1613836679&sr=8-1-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUExQzMzWkdMTDlBVldRJmVuY3J5cHRlZElkPUEwMTY1NTYxMUNKMk8zN1YwR0ZZTyZlbmNyeXB0ZWRBZElkPUEwMjIwMDEwMjRUUzk1MllXUU5DSSZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

The Supervised Learning Workshop: A New, Interactive Approach to Understanding Supervised Learning Algorithms

Sequence models are statistical supervised learning models that predict/classify based on sequence data.

Similar to other supervised statistical models, sequence models are in two types:
- Sequence prediction
- Sequence classification

Sequence Models

The purpose of a statistical learning model is to use a statistical model and the training data to find the function $\hat{f}$ as an estimate of the unknown function $f$ that maps the matrix of input variables $X$ to the outcome vector $Y$, as best as possible.
$$Y = f(X) + \epsilon \approx \hat{f}(X)$$ 

Purpose of supervised statistical learning

In supervised learning, input values represent the set of known features or attributes provided to a model to make predictions. A critical first step in data preprocessing is partitioning the dataset to separate these input columns from the designated target values that the model aims to estimate.

Input Values

In information retrieval, search ranking is a task that involves imposing a rank on a set of items. Instead of simply determining if an item is relevant to a user's query, the objective is to assign a score to each relevant candidate and display the top-rated elements most prominently, utilizing machine learning and behavioral models to obtain query-dependent relevance scores.

Search Ranking

Sequence learning encompasses machine learning problems that require a model to ingest sequences of inputs, emit sequences of outputs, or both. Unlike standard tasks that process a fixed number of inputs independently, sequence learning models maintain context across successive observations, making them essential for handling varying-length data such as video snippets, text, or patient histories over time.

Sequence Learning

In supervised learning, a target value (often called a label) represents the designated output or objective variable that a model is trained to predict. During data preprocessing, the dataset is typically partitioned to isolate the target column from the input values, enabling the model to learn the mapping from features to the desired prediction.

Target Values

In standard supervised learning, we commonly rely on the independent and identically distributed (IID) assumption. This principle dictates that both the training and test data are drawn independently from the exact same underlying probability distribution, denoted as $$P(X,Y)$$. Without this strong assumption that $$P(\cdot) = Q(\cdot)$$ (where $$Q$$ is the test distribution), it would be impossible to justify using patterns learned from the training data to make predictions on new test data.

Learn Before

Related

Learn After