Software Semester Exam

Algorithm

A step-by-step set of instructions used to solve a problem or complete a task.

Sequencing

Executing instructions in a specific order, one after another.

Selection

Making decisions in an algorithm using conditions (e.g., IF statements).

Iteration

Repeating a set of instructions until a condition is met or for a set number of times.

Pseudocode

A structured, human-readable way of describing an algorithm without using programming language syntax.

Linear Search

A search method that checks each item in a list one by one until the target is found.

Binary Search

A search method that repeatedly halves a sorted list to locate a target value efficiently

Comparing Search Algorithms

Evaluating search methods based on speed, efficiency, and requirements.

Traditional Programming

Humans create rules and data is processed according to those rules to produce outputs.

Supervised Learning

Learning from labelled data where the correct answers are known.

Unsupervised Learning

Finding patterns or structures in unlabelled data.

Reinforcement Learning

Learning through rewards and penalties based on actions taken in an environment.

Data Dimensions

The number of features present in a dataset.

Trivial Features

Features with little usefulness or predictive value.

Non-Trivial Features

Features that provide meaningful information for making predictions.

Overfitting

When a model learns training data too closely and performs poorly on new data.

F1 Score

A metric that balances precision and recall into a single value.

Input Layer

The first layer that receives data into a neural network.

Hidden Layers

Intermediate layers that process and transform information.

Output Layer

The final layer that produces the network's prediction.

Neurons

Processing units that receive inputs and produce outputs.

Weights

Values that determine the importance of inputs in a neural network.

Biases

Additional values that help neurons adjust their output.

Activation Functions

Mathematical functions that determine whether a neuron activates.

Cost Functions

Functions that measure prediction error.

Learning Rate

A value that controls how much weights and biases are adjusted during learning.

Epochs

One complete pass through the entire training dataset.

Generalisation

A model's ability to perform well on unseen data.

Overfitting

When a neural network memorises training data instead of learning general patterns.

Imbalanced Dataset

A dataset where some classes have significantly more