Intro to AI(Week 2 - Section 1): Machine Learning Flashcards
(80 cards)
1
Q
Define: Artificial Intelligence (AI)
A
A field of computer science that focuses on creating intelligent machines capable of performing tasks that typically require human intelligence.
2
Q
Define: Machine Learning
A
A subset of AI that uses computer algorithms to analyze data and make intelligent decisions based on what it has learned.
3
Q
Define: Traditional Programming
A
The process of creating algorithms using data and rules to develop a program that provides specific answers.
4
Q
Define: Machine Learning Model
A
A set of rules and parameters created by a machine learning algorithm to determine the logic and predict values based on inputs.
5
Q
Define: Supervised Learning
A
A type of machine learning where an algorithm is trained on human-labeled data to classify new data accurately.
6
Q
Define: Unsupervised Learning
A
A type of machine learning where an algorithm finds patterns in unlabeled data without human-labeled guidance.
7
Q
Define: Reinforcement Learning
A
A type of machine learning where an algorithm learns how to achieve goals by trying different combinations of allowed actions and receiving rewards or punishments.
8
Q
Define: Behavioral Rules
A
Rules defined by examining and comparing large datasets to find common patterns in machine learning.
9
Q
Define: Clustering
A
A technique used in unsupervised learning to group similar data together based on their similarities.
10
Q
Define: Baseline
A
The normal or expected behavior of a system or dataset used as a reference point for comparison in machine learning.
11
Q
Define: Outlier
A
Data points that significantly differ from the normal or expected behavior in machine learning.
12
Q
Define: Malicious Behavior
A
Actions or patterns that indicate harmful or unauthorized activities in machine learning.
13
Q
Define: State
A
The current condition or situation in reinforcement learning that defines the starting point for decision-making.
14
Q
Define: Allowed Actions
A
Actions that a machine learning algorithm can take within defined constraints to achieve its goals in reinforcement learning.
15
Q
Define: Rewards and Punishments
A
Feedback given to a machine learning algorithm based on the quality of its decisions in reinforcement learning.
16
Q
What are the three categories of Machine Learning?
A
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning.
17
Q
What is Supervised Learning?
A
It refers to building a classification model using class labels in the dataset.
18
Q
What is Unsupervised Learning?
A
It involves discovering class labels from unstructured data without using class labels.
19
Q
What is Reinforcement Learning?
A
It uses a reward function to penalize bad actions or reward good actions.
20
Q
What are the three categories of Supervised Learning?
A
- Regression
- Classification
- Neural Networks
21
Q
What is Regression?
A
It estimates continuous values by looking at the relationships between features and the result.
22
Q
What are Neural Networks?
A
They imitate the structure of the human brain.
23
Q
What is Classification?
A
It focuses on discrete values and assigns class labels based on input features.
24
Q
What are some examples of Classification models?
A
Decision trees, support vector machines, logistic regression, and random forests.
25
What is the purpose of the Training, Validation, and Test sets in Machine Learning?
1. The Training set is used to train the algorithm.
2. The Validation set is used to validate and fine-tune the algorithm's parameters.
3. Test set is data the model has never seen before and used to evaluate the model's performance.
26
What is a Neural Network?
A Neural Network is a computational model inspired by the structure and functioning of the human brain.
27
What are the basic components of a Neural Network?
The basic components of a Neural Network are:
1. input layer
2. hidden layer(s)
3. output layer.
28
What is the purpose of the hidden layer(s) in a Neural Network?
The hidden layer(s) in a Neural Network perform complex computations and help in learning and extracting features from the input data.
29
What is the activation function in a Neural Network?
The activation function introduces non-linearity to the Neural Network, allowing it to learn and model complex relationships in the data.
30
What is backpropagation in Neural Networks?
Backpropagation is an algorithm used to train Neural Networks by adjusting the weights and biases based on the error between predicted and actual outputs.
31
What is the purpose of the input layer in a Neural Network?
The input layer in a Neural Network receives the initial input data and passes it to the hidden layer(s) for further processing.
32
What is the purpose of the output layer in a Neural Network?
The output layer in a Neural Network produces the final output or prediction based on the computations performed in the hidden layer(s).
33
How does information flow in a Neural Network?
In a Neural Network, information flows from the input layer to the hidden layer(s), where computations are performed, and then to the output layer for the final prediction or output.
34
What is classification in machine learning?
Classification is the process of predicting the class or category of given data points based on input features.
35
What are some examples of classification algorithms?
Examples of classification algorithms include decision trees, support vector machines, logistic regression, and random forests.
36
How do classification models extract features from data?
Classification models extract features from data by identifying distinctive properties of input patterns that help determine the output categories or classes.
37
What is the purpose of the training, validation, and test sets in machine learning?
The training set is used to train the algorithm, the validation set is used to fine-tune the algorithm's parameters and validate results, and the test set is used to evaluate the model's performance on unseen data.
38
How does the training process work in classification?
During training, a learning algorithm is used to determine and develop the parameters of the classification model by repeatedly showing it labeled data and modifying internal values until it learns to make accurate predictions.
39
What is a Support Vector Machine (SVM) in machine learning?
A support vector machine is a supervised learning algorithm used for classification and regression tasks. It works by finding the optimal hyperplane that separates different classes or predicts continuous values. SVMs maximize the margin between the classes, with support vectors being the data points closest to the decision boundary.
40
What is a decision tree in machine learning?
A decision tree is a supervised learning algorithm that uses a tree-like model to make decisions or predictions by mapping input features to output labels or values. It consists of nodes representing features, branches representing decisions, and leaves representing the final output or prediction.
41
What is a regression task in machine learning?
A regression task is a type of supervised learning problem where the goal is to predict a continuous output variable based on input features. Regression models estimate the relationship between the input features and the output variable, allowing for the prediction of numeric values rather than discrete classes.
42
What is logistic regression in machine learning?
Logistic regression is a supervised learning algorithm used for binary classification tasks. It models the relationship between input features and the probability of the output belonging to a specific class. Unlike linear regression, logistic regression uses a logistic function to map the input features to a probability value between 0 and 1. It is commonly used when the dependent variable is categorical or binary.
43
What is a random forest in machine learning?
A random forest is an ensemble learning method that combines multiple decision trees to make predictions. It is effective for both classification and regression tasks, and it helps to handle high-dimensional data and avoid overfitting.
44
Define Decision Trees (in Random Forest)
Random forest is an ensemble of decision trees. Decision trees are the building blocks of a random forest, where each tree is trained on a random subset of the training data.
45
Define Random Subset of Features (in Random Forest)
Random Forest selects a random subset of features from the input data for each decision tree. This helps to introduce diversity among the trees and prevents overfitting.
46
Define Bagging (in Random Forest)
Random forest uses a technique called bagging (bootstrap aggregating) to create multiple subsets of the training data. Each decision tree is trained on a different subset, allowing for a diverse set of trees.
47
Define Voting or Averaging (in Random Forest)
During prediction, random forest combines the predictions of all the decision trees. For classification tasks, it uses majority voting to determine the final class label. For regression tasks, it averages the predictions of all the trees to obtain the final output.
48
Define Ensemble Learning (in Random Forest)
Random forest is an ensemble learning method, meaning it combines the predictions of multiple models (decision trees) to make a final prediction. This ensemble approach helps to improve the overall accuracy and robustness of the model.
49
Define Feature Importance (in Random Forest)
Random forest provides a measure of feature importance. It calculates the contribution of each feature in the model's performance, allowing for insights into which features are most influential in making predictions.
50
What is Deep Learning?
Deep Learning is a specialized subset of Machine Learning that uses layered algorithms to create a Neural Network, enabling AI systems to continuously learn and improve results.
51
How does Deep Learning differ from traditional Machine Learning?
Deep Learning algorithms rely on multiple layers of processing units, allowing them to learn from unstructured data and improve performance as more data is fed into the system.
52
What is the purpose of training a Deep Learning model?
Deep Learning models are trained by providing them with annotated examples, such as images with corresponding labels, to detect common patterns and improve accuracy.
53
What are some applications of Deep Learning?
Deep Learning is used in various tasks, including image captioning, voice recognition, facial recognition, medical imaging, language translation, and driverless cars.
54
How does Deep Learning enable natural language understanding?
Deep Learning algorithms enable AI systems to work out the context and intent of conveyed information, allowing for natural language understanding capabilities
55
Why is it called "deep" learning?
Deep Learning refers to the many layers of processing units in the neural network, which process and pass information to the next layer, creating a deep structure.
56
How does Deep Learning improve as datasets grow?
Unlike traditional Machine Learning algorithms, Deep Learning algorithms continue to improve performance as more data is fed into the system.
57
What is the role of Deep Learning in driverless cars?
Deep Learning is one of the main components of driverless cars, enabling them to process and interpret visual information for autonomous navigation.
58
What types of data can Deep Learning algorithms learn from?
Deep Learning algorithms can learn from unstructured data such as photos, videos, and audio files.
59
How does Deep Learning contribute to image recognition?
Deep Learning algorithms can detect common patterns in images by adjusting the weights of variables in each layer, allowing for accurate image recognition.
60
What is an artificial neural network?
A collection of computing units called neurons that mimic the human brain's information processing.
61
How do neural networks learn?
Through a process called backpropagation, where training data is used to adjust the network's weights and reduce errors.
62
What is a layer in a neural network?
A collection of neurons that takes in an input and provides an output.
63
What are hidden layers in a neural network?
Layers that simulate the activity in the human brain and produce an output through an activation function.
64
What is a deep neural network?
A neural network with more than one hidden layer.
65
What are perceptron’s?
The simplest type of neural networks consisting of input nodes connected directly to an output node.
66
What is the role of bias in neural networks?
Bias is a special type of weight that applies to a node after considering other inputs.
67
What is an activation function in a neural network? It determines how a node responds to its inputs and produces an output.
It determines how a node responds to its inputs and produces an output.
68
What are Convolutional Neural Networks (CNNs) used for?
They are useful in applications such as image processing, video recognition, and natural language processing.
69
What are Recurrent Neural Networks (RNNs)?
RNNs perform the same task for every element of a sequence, with prior outputs feeding subsequent stage inputs.
70
What is a Convolutional Neural Network (CNN)?
A Convolutional Neural Network (CNN) is a type of multilayer neural network that takes inspiration from the animal visual cortex. CNNs are particularly useful in applications such as image processing, video recognition, and natural language processing. They are adept at detecting simple structures in an image and constructing more complex features by combining these simple features through a series of convolutional layers.
71
What are Recurrent Neural Networks (RNNs)?
Recurrent Neural Networks (RNNs) are a type of neural network that perform the same task for every element of a sequence, with prior outputs feeding subsequent stage inputs. Unlike traditional neural networks, RNNs can make use of information in long sequences by representing each layer of the network as an observation at a certain time. RNNs are particularly useful in scenarios where the context or dependence on previous observations is important for producing accurate outputs, such as natural language processing, speech recognition, and time series analysis.
72
What task is a Recurrent Neural Network (RNN) best suited for?
Recurrent Neural Networks (RNNs) are best suited for tasks that involve sequential data or time series analysis. RNNs excel in scenarios where the context or dependence on previous observations is important for producing accurate outputs. They are commonly used in natural language processing, speech recognition, machine translation, sentiment analysis, and handwriting recognition, among other applications. RNNs can effectively capture the temporal dependencies in sequential data, making them a powerful tool for tasks that involve analyzing and generating sequences.
73
What task is a Convolutional Neural Network (CNN) best suited for?
Convolutional Neural Networks (CNNs) are best suited for tasks that involve image processing, video recognition, and natural language processing. CNNs are particularly effective in detecting simple structures in an image and constructing more complex features by combining these simple features through a series of convolutional layers. They excel at tasks such as image classification, object detection, facial recognition, and scene understanding. CNNs have revolutionized computer vision and are widely used in various applications that involve visual data analysis.
74
What is a neuron (in a neural network)?
A neuron is a fundamental unit of a neural network, inspired by the biological neurons in the human brain. It is a computational unit that receives input signals, processes them, and produces an output signal. Neurons are connected to each other through weighted connections, and they use an activation function to determine the output based on the weighted sum of inputs. Neurons play a crucial role in information processing and decision-making within neural networks.
75
What is backpropagation?
Backpropagation is a learning algorithm used in artificial neural networks to train the network by adjusting the weights of the connections between neurons. It involves a two-step process: forward propagation and backward propagation. During forward propagation, the input data is fed into the network, and the output is calculated layer by layer until the final output is obtained. Then, during backward propagation, the error between the predicted output and the desired output is calculated. This error is then propagated backward through the network, adjusting the weights of the connections in a way that reduces the error. The process is repeated iteratively until the network learns to produce accurate outputs for the given inputs. Backpropagation is a key component in training neural networks and allows them to learn from labeled training data.
76
What are the steps involved in backpropagation?
Backpropagation involves the following steps:
1. Forward Propagation: The input data is fed into the neural network, and the output is calculated layer by layer until the final output is obtained.
2. Calculate Error: The error between the predicted output and the desired output is calculated using an error function. This function determines how far the given output is from the desired output.
3. Backward Propagation: The error is propagated backward through the network. Starting from the output layer, the error is used to adjust the weights of the connections between neurons.
4. Update Weights: The weights are updated by applying an optimization algorithm, such as gradient descent. This algorithm adjusts the weights in a way that reduces the error.
5. Repeat: Steps 1 to 4 are repeated iteratively for a set number of epochs or until the network converges to a satisfactory level of accuracy.
77
What is forward propagation?
Forward propagation, also known as feedforward, is the process in which input data is fed into a neural network, and the output is calculated layer by layer until the final output is obtained. During forward propagation, the input values are multiplied by the weights of the connections between neurons and passed through an activation function. This activation function introduces non-linearity into the network, allowing it to learn complex patterns and make predictions. The output of one layer serves as the input for the next layer until the final output layer is reached. Forward propagation is the initial step in the neural network's computation and is used to generate predictions or outputs based on the given input data.
78
What are the steps involved in forward propagation?
Forward propagation involves the following steps:
1. Input Layer: The input data is fed into the neural network through the input layer. Each input node represents a feature or attribute of the data.
2. Weighted Sum: The input values are multiplied by the weights of the connections between neurons in the next layer. These weighted inputs are summed up.
3. Activation Function: The sum of the weighted inputs is passed through an activation function, which introduces non-linearity into the network. Common activation functions include sigmoid, ReLU, and tanh.
4. Output Calculation: The output of the activation function becomes the input for the next layer. This process is repeated for each layer until the final output layer is reached.
5. Final Output: The final output layer produces the predicted output or classification based on the input data and the learned weights.
79
What are Perceptrons?
Perceptrons are the simplest and oldest types of neural networks. They are single-layered neural networks consisting of input nodes connected directly to an output node. Perceptrons receive input values from the input layer and forward them to the next layer by multiplying them with weights and summing the results. Each input value is associated with a weight, which determines its importance in the network's decision-making process. The weighted inputs are then summed up, and an activation function is applied to the sum to determine the output of the perceptron. The activation function introduces non-linearity and helps the perceptron make decisions based on the input data. Perceptrons are often used for binary classification tasks, where they learn to classify input data into two categories based on the learned weights and activation function.
80
What are the steps a perceptron takes?
A perceptron follows these steps:
1. Input: The perceptron receives input values from the input layer. Each input value is associated with a weight, which determines its importance in the decision-making process.
2. Weighted Sum: The input values are multiplied by their corresponding weights. These weighted inputs are then summed up.
3. Activation Function: The sum of the weighted inputs is passed through an activation function. The activation function introduces non-linearity and determines the output of the perceptron.
4. Output: The output of the activation function becomes the output of the perceptron. It can be a binary output (0 or 1) or a continuous value, depending on the problem being solved.
