Finals Lectures Flashcards
1
Q
Identify
System of interconnected neurons that mimics the human brain’s ability to process information
2 words
NEURAL NETWORK
A
Neural Network
2
Q
Identify
Can recognize patterns and correlations in raw data, cluster and classify it, and continuously learn and improve.
2 words
NEURAL NETWORK
A
Neural Network
3
Q
Identify
Basic building block of a neural network, modeled after biological neurons in the human brain.
1 word
NEURAL NETWORK
A
Neuron
4
Q
Identify
Receives inputs, processes them by applying weights and biases, and passes the result through an activation function to produce an output.
1 word
NEURAL NETWORK
A
Neuron (Function)
5
Q
Identify
Parameters in a neural network that determine the importance of each input.
1 word
NEURAL NETWORK
A
Weights
6
Q
Identify
They scale the input values before they are summed up in the neuron. Inputs with higher weights contribute more to the neuron’s decision.
1 word
NEURAL NETWORK
A
Weights (Purpose)
7
Q
Identify
Constant value added to the weighted sum of inputs before applying the activation function.
1 word
NEURAL NETWORK
A
Bias
8
Q
Identify
It helps shift the activation function curve to better fit the data, allowing the model to represent patternsmore flexibly.
1 word
NEURAL NETWORK
A
Bias (Purpose)
9
Q
Identify
Determines whether a neuron ‘fires’ and produces an output.
2 words
NEURAL NETWORK
A
Activation Function
10
Q
Identify
It introduces non-linearity into the network, enabling it to learn complex patterns.
2 words
NEURAL NETWORK
A
Activation Function
11
Q
Identify
Produces outputs between 0 and 1, useful for probabilities.
2 words
NEURAL NETWORK - Common Activation Function
A
Sigmoid Function
12
Q
Identify
Outputs the input directly if it’s positive; otherwise, 0.
3 words or shortened 1 word
NEURAL NETWORK - Common Activation Function
A
ReLU (Rectified Linear Unit)
13
Q
Identify
Training, validation, and test sets.
1 word
NEURAL NETWORK - Common Activation Function
A
Dataset
14
Q
Identify
Data flows through the network.
2 words
NEURAL NETWORK - Common Activation Function
A
Forward Propagation
15
Q
Identify
Adjusts weights using the error gradient.
2 words
NEURAL NETWORK - Common Activation Function
A
Backward Propagation
16
Q
Identify
Repeated adjustments to improve performance.
1 word
NEURAL NETWORK - Common Activation Function
A
Iterations (Epochs)
17
Q
Identify
Simplest type of neural network.
3 words or abbreviation
NEURAL NETWORK - Types of Neural Network
A
Feedforward Neural Network (FNN)
18
Q
Identify
information flows in one direction—from input to output
3 words or abbreviation
NEURAL NETWORK - Types of Neural Network
A
Feedforward Neural Network (FNN)
19
Q
Identify
Used for tasks where inputs and outputs are straightforward.
3 words or abbreviation
NEURAL NETWORK - Types of Neural Network
A
Feedforward Neural Network (FNN)
20
Q
Identify
No complex patterns or sequences are involved; it just maps input data to output values.
3 words or abbreviation
NEURAL NETWORK - Types of Neural Network
A
Feedforward Neural Network (FNN)
21
Q
Identify (Example)
Predicting house prices
1 word
NEURAL NETWORK - Types of Neural Network
A
Task (Feedforward Neural Network (FNN))
22
Q
Identify (Example)
House size, number of bedrooms, location rating.
1 word
NEURAL NETWORK - Types of Neural Network
A
Inputs (Feedforward Neural Network (FNN))
23
Q
Identify (Example)
Estimated price of the house.
1 word
NEURAL NETWORK - Types of Neural Network
A
Output (Feedforward Neural Network (FNN))
24
Q
Identify
Specially designed for image and video data.
3 words or abbreviation
NEURAL NETWORK - Types of Neural Network
A
Convolutional Neural Networks (CNN)
25
# Identify
**Recognizes patterns**, edges, and features in visuals by **applying filters**.
| 3 words or abbreviation
## Footnote
NEURAL NETWORK - Types of Neural Network
Convolutional Neural Networks (CNN)
26
# Identify
**It’s excellent at analyzing spatial features like shapes**, textures, and colors in images.
| 3 words or abbreviation
## Footnote
NEURAL NETWORK - Types of Neural Network
Convolutional Neural Networks (CNN)
27
# Identify (Example)
Identifying cats and dogs in pictures.
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Task (Convolutional Neural Networks (CNN))
28
# Identify (Example)
Images of animals
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Inputs (Convolutional Neural Networks (CNN))
29
# Identify (Example)
Label ("Cat" or "Dog")
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Output (Convolutional Neural Networks (CNN))
30
# Identify
**Processes sequential data by remembering past inputs** while processing current ones
| 3 words or abbreviation
## Footnote
NEURAL NETWORK - Types of Neural Network
Recurrent Neural Network (RNN)
31
# Identify
**Useful for tasks** where the **order of data matters**
| 3 words or abbreviation
## Footnote
NEURAL NETWORK - Types of Neural Network
Recurrent Neural Network (RNN)
32
# Identify
**It can retain information about previous words** to make the prediction contextually accurate
| 3 words or abbreviation
## Footnote
NEURAL NETWORK - Types of Neural Network
Recurrent Neural Network (RNN)
33
# Identify (Example)
Predicting the next word in a sentence.
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Task (Predicting the next word in a sentence.)
34
# Identify (Example)
A sequence of words like "I love neural..."
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Inputs (Predicting the next word in a sentence.)
35
# Identify (Example)
Predicted word ("networks").
| 1 word
## Footnote
NEURAL NETWORK - Types of Neural Network
Output (Predicting the next word in a sentence.)
36
# Identify
**Process of discovering patterns** in **large datasets**.
| 2 words
## Footnote
Data Mining
Data Mining
37
# Identify
**It's a multidisciplinary field** that uses techniques from **machine learning, statistics and database systems.**
| 2 words
## Footnote
Data Mining
Data Mining
38
# Fill in the blanks
Extract **?** information
| 1 word
## Footnote
Data Mining - Goals
useful
39
# Fill in the blanks
Make **?** decisions
| 2 words
## Footnote
Data Mining - Goals
data-driven
40
# Identify
**Organized and easy to search** and analyze because it is **stored in predefined formats.**
| 2 words
## Footnote
Data Mining - Types of Data
Structured Data
41
# Identify
**It has rows and columns**, with each column **assigned a specific type of data.**
| 2 words
## Footnote
Data Mining - Types of Data
Structured Data
42
# Identify
**Does not have a fixed format**, making it **harder to organize and analyze directly.**
| 2 words
## Footnote
Data Mining - Types of Data
Unstructured Data
43
# Identify
**It can include text, images, audio, video**, or any other data format that **doesn't fit neatly into rows and columns.**
| 2 words
## Footnote
Data Mining - Types of Data
Unstructured Data
44
# Identify
**The crucial first step in any data analysis** or machine learning pipeline.
| 2 words
## Footnote
Data Mining
Data Preprocessing
45
# Identify
**It involves cleaning, organizing, and transforming raw** data into a usable format for analysis.
| 2 words
## Footnote
Data Mining
Data Preprocessing
46
# Identify
**?** is often **incomplete, noisy, or inconsistent, making it difficult to analyze** or use effectively.
| 2 words
## Footnote
Data Mining - Data Preprocessing
Raw data
47
# Identify
**?** ensures the quality, reliability, and performance of your models.
| 2 words
## Footnote
Data Mining - Data Preprocessing
Proper preprocessing
48
# Identify
**replace missing values with mean/median** (for numerical data) or mode (for categorical data), or remove rows/columns with too many missing values.
| 2 words and 3 words (Key Step Term and Term under it)
## Footnote
Data Mining - Key Steps in Data Preprocessing
Data Cleaning (Handling Missing Values)
49
# Identify
**Eliminate irrelevant data points**, outliers, or duplicates.
| 2 words and 2 words (Key Step Term and Term under it)
## Footnote
Data Mining - Key Steps in Data Preprocessing
Data Cleaning (Removing Noise)
50
# Identify
**Combine data** from multiple sources
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing
Data Integration
51
# Identify (Example)
**Combine data** from multiple sources
| 1 word
## Footnote
Data Mining - Key Steps in Data Preprocessing
Tasks (Data Integration)
52
# Identify
**Convert data** into a format **suitable for analysis or modeling.**
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing
Data Transformation (Objective )
53
# Identify
**Scale numerical data** to a **common range (e.g., 0 to 1).**
| 1 word
## Footnote
Data Mining - Key Steps in Data Preprocessing - Objective Tasks
Normalization
54
# Identify
**Convert continuous data** **into discrete categories** (e.g., age groups).
| 1 word
## Footnote
Data Mining - Key Steps in Data Preprocessing - Objective Tasks
Discretization
55
# Identify
**Reduce the size of data** while maintaining essential information
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing
Data Reduction (Objective)
56
# Identify
**Use techniques like Principal Component Analysis (PCA)** to **reduce the number of features.**
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing - Data Reduction Tasks
Dimensionality Reduction
57
# Identify
**Use techniques like Principal Component Analysis (PCA)** to **reduce the number of features.**
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing - Data Reduction Tasks
Dimensionality Reduction
58
# Identify
**Select only the most relevant** attributes for analysis.
| 2 words
## Footnote
Data Mining - Key Steps in Data Preprocessing - Data Reduction Tasks
Feature Selection
59
# Identify
**Supervised machine learning technique** used to **predict the category or label of data points** based on input features.
| 1 word
## Footnote
Data Mining - Classification
Classification
60
# Identify
It **requires a training dataset** with **predefined labels**
| 1 word
## Footnote
Data Mining - Classification
Classification
61
# Enumerate (Examples)
Classification Examples
| Examples only
## Footnote
Data Mining - Classification
* Email spam detection
* Loan approval
62
# Enumerate (Examples)
Classification Techniques
| Examples only
## Footnote
Data Mining - Classification
* Decision Trees
* Naïve
* Bayes
* Support Vector Machines (SVM)
63
# Enumerate (Examples)
Classification Evaluation
| Examples only
## Footnote
Data Mining - Classification
* Accuracy
* Precision
* Recall
* F1-score
64
# Identify
**Unsupervised machine learning technique** that **groups similar data points** into clusters.
| 1 word
## Footnote
Data Mining - Clustering
Clustering
65
# Enumerate (Examples)
Clustering Examples
| Examples
## Footnote
Data Mining - Clustering
* Customer segmentation
* Image segmentation
66
# Enumerate (Examples)
Clustering Techniques
| Examples
## Footnote
Data Mining - Clustering
* K-Means
* Hierarchical
* Clustering
* DBSCAN
67
# Explain
Clustering Evaluation
| Examples
## Footnote
Data Mining - Clustering
Choosing the right number of clusters
68
# Identify
**Technique to discover interesting relationships** or patterns between items in a dataset.
| 3 words
## Footnote
Data Mining - Association Rule Mining
Association Rule Mining
69
# Identify
It is **commonly used in market basket analysis.**
| 3 words
## Footnote
Data Mining - Association Rule Mining
Association Rule Mining
70
# Enumerate (Examples)
Association Rule Mining (Examples)
| Examples
## Footnote
Data Mining - Association Rule Mining
* Market Basket Analysis
* Online Recommendations
71
# Identify
**Frequency of an itemset** in the dataset.
| 1 word
## Footnote
Data Mining - Association Rule Mining - Key concepts
Support
72
# Identify
**Strength of a rule** (e.g., if A, then B).
| 1 word
## Footnote
Data Mining - Association Rule Mining - Key concepts
Confidence
73
# Identify
**Measures how much more likely two items are bought together** than if they were independent.
| 1 word
## Footnote
Data Mining - Association Rule Mining - Key concepts
Lift
74
# Identify
**Machine learning technique** that uses **labeled data to train algorithms.**
| 2 words
## Footnote
Supervised Learning
Supervised Learning
75
# Identify
**?** are trained **using input objects and desired output values**, often human-made labels.
| 3 words
## Footnote
Supervised Learning
Supervised learning models
76
# Identify
**Can identify** and **classify spam emails**.
| 2 words
## Footnote
Supervised Learning - Applications
Spam filtering
77
# Identify
**Can automatically classify images** into different categories.
| 2 words
## Footnote
Supervised Learning - Applications
Image classification
78
# Identify
**Can analyze patient data** to **identify patterns** that **suggest specific diseases** or conditions.
| 2 words
## Footnote
Supervised Learning - Applications
Medical diagnosis
79
# Identify
**Can analyze financial transactions** to **identify patterns that indicate fraudulent activity**.
| 2 words
## Footnote
Supervised Learning - Applications
Fraud detection
80
# Identify
**Models can overfit training data**, leading to **poor performance on new data.**
| 1 word
## Footnote
Supervised Learning - Challenges
Overfitting
81
# Identify
**Bias in the training data** may result in **unfair predictions.**
| 3 words
## Footnote
Supervised Learning - Challenges
Bias in models
82
# Identify
**Labeled training data can be costly** and time-consuming to obtain.
| 4 words
## Footnote
Supervised Learning - Challenges
Dependence on labeled data
83
# Identify
**Missing data** can **disrupt model training**
| 3 words
## Footnote
Supervised Learning - Data Preprocessing
Handle Missing Values
84
# Identify
**Replacing missing values with the mean (for numerical data)** or mode (for categorical data) **maintains dataset integrity**
| 3 words
## Footnote
Supervised Learning - Data Preprocessing
Handle Missing Values
85
# Identify
**Supervised learning models require numerical inputs.**
| 3 words
## Footnote
Supervised Learning - Data Preprocessing
Encode Categorical Variables
86
# Identify
**Transforming categorical data** into **numeric form** allows the model to interpret them.
| 3 words
## Footnote
Supervised Learning - Data Preprocessing
Encode Categorical Variables
87
# Identify
**Different features often have varying magnitudes** (e.g., age in years vs. salary in thousands).
| 2 words
## Footnote
Supervised Learning - Data Preprocessing
Feature Scaling
88
# Identify
**Algorithms like Logistic Regression or SVM** are sensitive to these differences, as they **rely on distance or gradient-based optimization. **
| 2 words
## Footnote
Supervised Learning - Data Preprocessing
Feature Scaling
89
# Identify
**Scaling ensures equal importance** for all features
| 2 words
## Footnote
Supervised Learning - Data Preprocessing
Feature Scaling
90
# Identify
**Separates the dataset** into **training and testing subsets.**
| 2 words
## Footnote
Supervised Learning - Data Preprocessing
Data Splitting
91
# Identify
**?** is used to **build the model**, and **testing data evaluates its performance.**
| 2 words
## Footnote
Supervised Learning - Data Preprocessing - Data Splitting
Training data
92
# Identify
**Used for binary classification problems** (e.g., Yes/No, 0/1).
| 2 words
## Footnote
Supervised Learning - Logistic Regression
Logistic Regression
93
# Identify
**Estimates the probability of an observation belonging to a specific category** by **using the logistic function** (sigmoid function).
| 2 words
## Footnote
Supervised Learning - Logistic Regression
Logistic Regression
94
# Identify
The **?** **maps predictions to a range between 0 and 1**, making it interpretable as a probability.
| 2 words
## Footnote
Supervised Learning - Logistic Regression
sigmoid function
95
# Identify
**Powerful supervised learning algorithm** used **for classification and regression tasks.**
| 3 words
## Footnote
Supervised Learning - Support Vector Machine
Support Vector Machine
96
# Identify
In classification, **SVM aims to find the hyperplane** that best **separates the data into different classes.**
| 3 words
## Footnote
Supervised Learning - Support Vector Machine
Support Vector Machine
97
# Identify
The **?** is **the line (in 2D) or a plane (in 3D) that divides the feature space into regions** corresponding to different classes.
| 1 word
## Footnote
Supervised Learning - Support Vector Machine
hyperplane
98
# Identify
**The model seeks to maximize the margin**, which is the **distance between the hyperplane and the closest data points** from each class, called **?**.
| 2 words
## Footnote
Supervised Learning - Support Vector Machine
support vectors
99
# Identify
**You have points in a feature space (2D or higher)** where **each point belongs to a certain class.**
| 2 words
## Footnote
Supervised Learning - Support Vector Machine - Key Components
Data Points
100
# Identify
**Each point belongs to one of the categories** (like Class A or Class B).
| 1 word
## Footnote
Supervised Learning - Support Vector Machine - Key Components
Classes
101
# Identify
**SVM finds a line (in 2D) or plane (in higher dimensions)** that separates the data points into different classes.
| 1 word
## Footnote
Supervised Learning - Support Vector Machine - Key Components
Hyperplane
102
# Identify
**The points closest to the decision boundary (line)** that help determine the position of the hyperplane.
| 2 words
## Footnote
Supervised Learning - Support Vector Machine - Key Components
Support Vectors
103
# Fill in the blanks
The SVM model is **?** using a dataset.
| 1 word
## Footnote
Supervised Learning - Support Vector Machine - Steps - Training
trained
104
# Fill in the blanks
**Once trained**, the SVM can **?** **which side of the line a new data point falls into**—whether it belongs to Class A or Class B.
| 1 word
## Footnote
Supervised Learning - Support Vector Machine - Steps - Evaluation
predict
105
# Identify
Works by splitting the dataset into subsets based on feature values.
| 2 words
## Footnote
Supervised Learning - Decision Tree
Decision Tree
106
# Identify
**The process continues until the model achieves a desired level of purity** (where all samples in a subset belongs to the same class) or reaches a stopping criterion (e.g., maximum depth).
| 2 words
## Footnote
Supervised Learning - Decision Tree
Decision Tree
107
# Identify
**The dataset is split based on conditions**, like "Is Age > 30?" or "Is Salary > 50,000?".
| 1 word
## Footnote
Supervised Learning - Decision Tree - How it Works
Splitting
108
# Identify
**?** are made to **maximize information gain** (classification) or **minimize variance** (regression).
| 1 word
## Footnote
Supervised Learning - Decision Tree - How it Works
Splitting
109
# Identify
Each split leads to a **?**. **Nodes represent questions based on features.**
| 2 words
## Footnote
Supervised Learning - Decision Tree - How it Works
Decision Nodes
110
# Identify
**These are the final nodes that predict a class label** (for classification) or a continuous value (for regression).
| 2 words
## Footnote
Supervised Learning - Decision Tree - How it Works
Leaf Nodes
111
# Identify
**Determines how splits are made** (e.g., "gini" for Gini Impurity or "entropy" for Information Gain).
| 1 word
## Footnote
Supervised Learning - Decision Tree - How it Works - Key Parameters
Criterion
112
# Identify
**Limits the depth of the tree** to avoid overfitting.
| 2 words
## Footnote
Supervised Learning - Decision Tree - How it Works - Key Parameters
Max Depth
113
# Identify
**Minimum samples** required to **split a node.**
| 3 words
## Footnote
Supervised Learning - Decision Tree - How it Works - Key Parameters
Min Samples Split
114
# Identify
**Measures overall correctness** by **comparing correct predictions** to total predictions.
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Accuracy
115
# Identify
Useful when **class distributions are balanced**.
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Accuracy
116
# Identify
**Calculates the proportion of true positives** among predicted positives.
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Precision
117
# Identify
**Prioritize this** when **false positives are more costly** (e.g., spam detection).
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Precision
118
# Identify
**Measures sensitivity or the ability to identify** all actual positives.
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Recall
119
# Identify
Useful when **missing positives has severe consequences** (e.g., medical diagnosis).
| 1 word
## Footnote
Supervised Learning - Evaluating Models - Metrics
Recall
120
# Identify
**Provides a balance between precision and recall**, especially when classes are imbalanced.
| 2 words
## Footnote
Supervised Learning - Evaluating Models - Metrics
F1-Score
121
# Identify
**A simple way to evaluate how well a classification model works** by showing the actual and predicted results in a table.
| 2 words
## Footnote
Supervised Learning - Confusion Matrix
Confusion Matrix
