WorksheetsMachine Learning part 1
Total questions: 15
Worksheet time: 8mins
What is the main goal of machine learning?
To manually program all decision rules
To enable computers to learn from data
To increase hardware performance
To collect raw data only
How does deep learning differ from traditional machine learning?
It uses pre-defined rules
It uses multi-layer neural networks to learn features automatically
It does not require training data
It cannot process large data
Big Data is important in ML because it helps to:
Reduce storage requirements
Provide large-scale data for model training
Remove the need for preprocessing
Avoid computation
In the context of datasets, features are:
Output labels
Input variables or attributes describing samples
File formats
Data errors
Data preprocessing primarily involves:
Randomizing the dataset
Cleaning and normalizing data for better model performance
Removing all numeric columns
Reducing the number of algorithms
Which tool is primarily used for classical machine learning tasks?
WEKA
TensorFlow
Hadoop
AWS
Which of the following is not a machine learning library?
Pandas
NumPy
PowerPoint
Sklearn
The difference between supervised and unsupervised learning is that:
Supervised learning uses labeled data, while unsupervised does not
Unsupervised learning is slower
Both use the same data type
Only supervised learning can cluster data
What does the term sample refer to in a dataset?
The entire dataset
A single observation or data instance
The model parameters
A column header
In cybersecurity, datasets like NSL-KDD and UNSW-NB15 are used for:
Network intrusion detection and analysis
Sentiment analysis
Image recognition
Speech-to-text
In a dataset matrix, rows correspond to:
Features
Samples or observations
Labels
Hyperparameters
The code np.loadtxt(dataset, delimiter=",", skiprows=1) does what?
Loads an image file
Loads CSV data while skipping the header row
Visualizes data
Removes missing values
What are the three main dataset formats used in ML for cybersecurity?
.docx, .ppt, .pdf
.csv, .arff, .libsvm
.zip, .rar, .tar
.exe, .dll, .bat
The main difference between regular data science and cybersecurity data science is:
The algorithms used
The data type and how it is represented
The tools available
The size of datasets only
In a matrix representation of data, columns represent:
Samples
Features or attributes
Labels only
Model parameters
