Notebooks are computational documents used in Data Analysis, Data Science, and general Scientific Computing that allow for textual explanation and documentation to be presented alongside code (usually Python code).

What are Notebooks in Data Science?

Notebook

The command line is a text-based interface used to interact with a computer's operating system or command-line interface (CLI). It allows users to issue commands and perform various tasks by typing in text commands, rather than using a graphical user interface (GUI).

Command Line

In Python, the assignment operator is represented by the = symbol, which is used to assign a value to a variable. The basic syntax of the assignment operator is: variable_name = value

Assignment Operator Python

Assignment Operator (Python)

In Python, an "operator" is a symbol or a special set of characters that represents a specific action or operation to be performed on one or more operands (values or variables).

Operator (Python)

In Python, a variable is a named location in memory that stores a value. It can hold different types of data such as numbers, strings, lists, tuples, or even complex objects.

Variable (Python)

Data Science is an interdisciplinary field that involves the use of statistical, computational, and machine learning techniques to extract insights and knowledge from data. It combines elements from various fields such as statistics, mathematics, computer science, and domain expertise.

What is Data Science?

Data Science

The print function in Python is used to output text, variables, or expressions to the console or standard output device (like the output section of a code cell in a notebook).

Print Function (Python)

Snake Case is a naming convention (for naming variables, functions, files, etc). Where all letters are lowercase and words are separated by underscores. For example: snake_case_example

Snake Case (Python)

Snake Case

Artificial Intelligence (AI) is a branch of computer science that involves creating intelligent machines that can perform tasks that –used to– require human intelligence to accomplish. It encompasses several subfields, including: machine learning, natural language processing, computer vision, and robotics.

Artificial Intelligence

Artificial Narrow Intelligence (ANI) is a term used to describe artificial intelligence (AI) systems that are designed and programmed to perform a specific task or set of tasks.

Artificial Narrow Intelligence (ANI)

Artificial General Intelligence (AGI) refers to artificial intelligence that exhibits cognitive abilities comparable to humans across a wide range of domains. This is in contrast so-called "narrow" AI that may exhibit human-level capabilities, but only at specific tasks or in highly controlled environments.

Artificial General Intelligence (AGI)

A Data Scientist's work often involves applying machine learning algorithms and statistical models to solve business problems, make predictions, or uncover patterns within data.

Data Scientist

A data analyst is a professional who collects, cleans, and analyzes data to help businesses make better decisions. They extract meaning from data and communicate their findings to business owners, corporate decision-makers, and other stakeholders.

Data Analyst

A machine learning engineer specializes in designing, building, and deploying machine learning models and systems. This role combines elements of software engineering, data science, and applied mathematics.

Machine Learning Engineer (MLE)

Data Analysis is the exhilarating process of extracting meaning from piles of seemingly arbitrary numbers and turning them into information that is useful for decision making.

Data Analysis

Python is a programming language that emphasizes readability, simplicity, and ease of use. Python has gained widespread popularity in data science, and artificial intelligence.

Python Programming Language

Python

Exploratory Data Analysis (EDA)  involves the exploration, visualization, and summary of data to understand its main characteristics, identify patterns, detect anomalies, and test initial hypotheses.

Exploratory Data Analysis

Exploratory Data Analysis (EDA)

Business Intelligence (BI) refers to the set of strategies, tools, technologies, and processes used to collect, analyze, and present data in a format that helps businesses make informed decisions. BI primarily focuses on querying, reporting, data visualization, and data analytics to provide historical, current, and predictive views of business operations.

Business Intelligence (BI)

A False Negative represents an error that occurs when a test result or a classification model wrongly indicates a negative outcome when the actual situation is positive.

False Negative

A False Positive refers to an error that occurs when a test result or a classification model wrongly indicates a positive outcome when the actual situation is negative.

False Positive

Gradient Descent is an iterative optimization algorithm used to minimize the cost function or loss function in machine learning and optimization problems. It's how neural networks "learn".

Gradient Descent

Hyperparameters are like settings for the machine learning algorithm that are set before the training process begins. They typically remain constant while the model is trained.

Hyperparameter

In data science, imputation refers to the process of replacing missing or null values in a dataset with substituted values. Missing data is a common and can occur due to various reasons.

Imputation

K-means clustering is an unsupervised learning algorithm for assigning data points to one of a predetermined number of clusters based on the proximity of points to the cluster's average location or "centroid".

K-Means Clustering

Linear regression finds the best-fitting linear equation that describes the relationship between a dependent variable  "y" and one or more independent variables "x".

Linear Regression

Logistic regression is used primarily for binary classification, where the goal is to predict the probability of a binary outcome or event (such as Yes/No, True/False, 1/0).

Logistic Regression

A quantile is a statistical measure that divides a dataset or probability distribution into equal-sized parts, expressing the value below which a certain proportion of the data falls.

Quantile

A variable's quartiles are a set of three values that divides it into four equal parts.  The second of the three quartiles (Q2) is also known as the "median".

Quartile

A Random Forest is an ensemble of many different decision trees which –on aggregate– are able to make more accurate and robust predictions than any single decision tree.

Random Forest

The "target" is what a predictive model aims to predict or estimate based on the input data –also called the dependent variable, response variable, outcome variable, or label.

Target

Underfitting occurrs when a model is too simple to learn from the data effectively, resulting in poor performance both on the training data and unseen or "test" data.

Underfitting

In machine learning, a "weight" refers to a parameter used by the model to adjust the contribution of a feature or an input to the final output of the model.

Weight (Machine Learning)

Weight

In statistics, a Z-score (also known as a standard score) is a measure that expresses the number of standard deviations a data point is from the mean of a dataset.

Z-Score

An "x variable" typically refers to the independent variables or features in a dataset that are used to predict or estimate the dependent variable or target variable – "y."

X Variable

In machine learning and predictive modeling, the "y variable" is the thing that you want to predict. Also sometimes called the dependent variable or target variable.

Y Variable

Google Colaboratory, also known as Colab, is a free cloud-based platform provided by Google that allows users to write and execute Python code in a browser environment.

What is Google Colaboratory?

Google Colaboratory

An Artificial Neural Network (ANN) is a computational model inspired by the structure and functioning of the human brain's biological neural networks. It's a fundamental concept in machine learning and artificial intelligence.

Artificial Neural Network (ANN)

An algorithm is a step-by-step procedure designed to accomplish a particular task. Typically these unambiguous instructions are for computers or a humans to follow to perform a task or solve a problem.

Algorithm in Data Science

Algorithm

An activation function is a math function that determines the output of a neuron in a neural network. Common activation functions include: sigmoid, relu, tanh, and softmax.

Activation Function in Neural Networks

Activation Function

The accuracy score is a metric used to evaluate classification models. It is essentially the percentage of correct predictions made by the model: Accuracy = (Correct Predictions / Total Predictions) * 100

Accuracy Score in Classification

Accuracy Score

Big data refers to extremely large and complex datasets that cannot be easily managed, processed, or analyzed with traditional data processing tools. The term "big data" encompasses not only the size of the data but also its variety, velocity, and complexity.

What is Big Data?

Big Data

Business Analysts work to understand the business processes, identify areas for improvement, and propose solutions to meet business objectives. They serve as a liaison between different stakeholders, including business leaders, clients, and IT teams, to ensure that projects align with business goals.

Business Analyst (BA)

Business Analytics and Data Science are related fields that involve extracting insights from data to support decision-making processes, but they have distinct focuses and methodologies.

Business Analytics (BA)

Backpropagation is an algorithm used for training neural networks. The goal is to minimize the error of the network's output by adjusting the weights of the connections between neurons.

Backpropagation Algorithm - Neural Networks

Backpropagation

In the contexts of statistics and machine learning, "bias" refers to the systematic error or deviation of a model's predictions or estimates from the true values.

Bias in Machine Learning

Bias

The Bias-Variance Tradeoff deals with the balance between two types of errors: bias and variance –with the goal of developing models that generalize well to new, unseen data.

Bias-Variance Tradeoff

Unlike discrete variables, which can only take distinct, separate values, continuous variables can assume an infinite number of values within a specific interval. They can take on any real value within a given range.

Continuous Variable

A categorical variable is a type of variable in statistics that represents categories or groups. They can be further divided into two main types: nominal and ordinal.

Categorical Variable

Clustering in machine learning is a technique used to group similar data points together based on certain characteristics or features. The goal of clustering is to identify natural groupings in the data without any prior knowledge of the group memberships.

Clustering

Computer Vision is a field within data science and computer science that focuses on enabling machines to interpret and make decisions based on visual data.

Computer Vision

A cost function, also known as a loss function or objective function, quantifies the difference between the predicted values of a model and the actual values in the training data.

Cost Function

Cross-validation is a resampling technique used in machine learning to assess the performance and generalizability of a model. It helps ensure that the model's performance metrics are not overly dependent on the specific random partitioning of the data.

Cross-Validation

In statistics, variance measures the spread or dispersion of a set of data points. It quantifies how much the data points deviate from the mean (average) of the dataset.

Variance (statistics)

Variance

Word embeddings capture semantic relationships and meaning between words and represent them as a predetermined quantity of numbers held in a vector.

Word Embedding

In machine learning, "validation" is the process of evaluating a model's performance on a data that it was not trained on, to assess its ability to generalize to new, unseen data.

Validation (Machine Learning)

Validation

Validation Loss is a metric used to evaluate the performance of a trained model on a validation set. It is computed by comparing the model's predictions to the true labels.

Validation Loss

A validation set is a subset of the labeled data for assessing the performance of a model during the training phase. It independently evaluates the model's ability to generalize.

Validation Set

The Vanishing Gradient Problem occurs when the gradients of the loss function with respect to the weights become very small, causing the weights to be updated very slowly or not at all.

Vanishing Gradient Problem

Variable importances, also known as feature importances or attribute importances, refer to a measure of the contribution of individual variables in a  model to its performance.

Variable Importances

A Variational Autoencoder (VAE) is a type of generative model. VAEs belong to the family of autoencoders but incorporate probabilistic and variational methods to model complex distributions in the latent space.

Variational Autoencoder (VAE)

Unsupervised Learning is a category of machine learning algorithm used when a dataset does not include labelled examples –otherwise known as the "target variable"– within it.

Unsupervised Learning

Undersampling may be used to reduce the number of instances of the majority class to balance out the number of observations in the majority and minority classes.

Undersampling

Unlike structured data (which is typically stored in databases with a well-defined schema), unstructured data does not have a predefined data model or is not organized in a pre-defined manner.

Unstructured Data

Univariate analysis focuses on examining the distribution and characteristics of one variable in isolation, without considering the relationships between multiple variables.

Univariate Analysis

Tableau is widely used data visualization and business intelligence (BI) software that allows users to connect, visualize, and share data to facilitate understanding and decision-making.

Tableau

A sequence of data points collected or recorded over a period of time, typically at equally spaced intervals –common in domains like finance, economics, meteorology, etc.

Time Series

Time series data is a type of data where observations or measurements are collected and recorded over time, with each data point associated with a specific timestamp.

Time Series Data

A train-test split is common in machine learning for evaluating the performance of a model. Data is divided into two subsets: for training (train), and for evaluating (test).

Train Test Split

The t-distribution is a probability distribution for estimating population parameters when the sample size is small and the population standard deviation is unknown.

T-distribution

A t-test is a statistical test used to compare the means of two groups and determine whether there is a significant difference between them.

T-test

In the context of machine learning and deep learning, a tensor is a multi-dimensional array or a mathematical object that generalizes the concept of scalars, vectors, and matrices.

Tensor

A Tensor Processing Unit (TPU) is a specialized hardware accelerator designed by Google for machine learning workloads, particularly those associated with deep neural networks.

Tensor Processing Unit (TPU)

Type I error, also known as a "false positive" or "alpha error," occurs in statistical hypothesis testing when a true null hypothesis is incorrectly rejected.

Type I Error

Type II error, also known as a "false negative" or "beta error," occurs in statistical hypothesis testing when a false null hypothesis is not rejected.

Type II Error

Early stopping is a regularization technique used in training neural networks. The goal of early stopping is to prevent overfitting and improve the generalization of the model.

Early Stopping

A Hidden Layer refers to any layer of nodes between the input layer and the output layer in a neural network. They enable the model to represent complex data relationships.

Hidden Layer

In k-fold cross validation the dataset is split into multiple subsets or "folds" and then systematically each using each subset as both a training set and a validation set.

K-fold Cross Validation

In statistics, the "mean" is a measure of central tendency AKA the "average" of a set of values. It is calculated by summing up all values and then dividing by the number of observations.

Mean

A neural network is a type of machine learning algorithm consisting of interconnected layers of nodes or "neurons". It is said to mimic the learning processes of the human brain.

Neural Network

One-hot encoding is used to represent categorical variables as a set of binary vectors. It is common when dealing with categorical features in machine learning models.

One-Hot Encoding

Recall, also known as sensitivity or the true positive rate, calculates the ratio of true positive predictions to the total number of actual positive instances.

Recall

A scalar refers to a single numerical value, which can be a real number or a complex number. Scalars are used to scale vectors, matrices, and other linear transformations.

Scalar (Linear Algebra)

Scalar

Image recognition involves the identification and categorization of objects or patterns within images. The goal is to develop models that can automatically label images based on their content.

Image Recognition

TensorFlow is an open-source machine learning framework. It is widely used for building and training machine learning models –particularly neural networks.

What is TensorFlow?

TensorFlow

TensorBoard is a web-based visualization tool that comes with TensorFlow. It helps users understand, optimize, and debug machine learning models during the training process.

TensorBoard

In machine learning, the term "test loss" typically refers to the performance of a trained model on a separate dataset that it has not seen during the training phase.

Test Loss

In machine learning, a test set is a separate dataset used to assess the performance and generalization ability of a trained machine learning model.

Test Set

notebook meta title

Training Loss

A training set is a random subset of a dataset used to train a machine learning model. The model learns patterns and relationships within the data to help it make predictions.

Training Set

Tokenization is the process of breaking down a text into individual units, known as "tokens". These tokens could be words, subwords, or even characters (but are usually words).

Tokenization

Transfer learning is a machine learning technique where a model trained on one task is adapted to perform a different, but related, task. Pre-trained models have a number of advantages.

Transfer Learning

A true negative (TN) occurs when the model correctly identifies instances of the negative class. The model successfully recognizes cases where the true class is negative.

Data Science Encyclopedia

Use the encyclopedia to deepen your understanding of specific terms and topics.

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

S

T

U

V

W

X

Y

Z

A