Artificial Intelligence — Set 2

Correct Answer: D. Machine Learning

Deep Learning is a specific type of machine learning based on artificial neural networks with multiple layers. It is capable of learning unsupervised from data that is unstructured or unlabeled. Most high-end voice recognition systems use deep learning techniques.

Correct Answer: A. Errors caused by prejudiced assumptions in the data

AI bias occurs when an algorithm produces systematically prejudiced results due to flawed training data. It is a major ethical concern as it can lead to unfair treatment in hiring or law enforcement. Developers work to minimize bias to ensure AI systems are objective.

Correct Answer: C. A medical diagnosis program that uses rules

Expert systems are AI programs designed to solve complex problems by reasoning through bodies of knowledge. They use 'if-then' rules to mimic the decision-making ability of a human expert. They were among the first truly successful forms of AI software.

Correct Answer: D. To gather data about the physical environment

Sensors allow robots to 'see', 'hear', and 'feel' their surroundings to make intelligent decisions. Common sensors include cameras, LIDAR, and tactile pressure pads. They are the primary source of input data for autonomous physical systems.

Correct Answer: A. Finding hidden patterns or clusters in unlabeled data

Unsupervised learning looks for similarities in data that has no pre-existing labels. It is widely used for market segmentation and identifying similar customer groups. The algorithm organizes the data solely based on internal structures.

Correct Answer: C. Python

Python is preferred for AI due to its simplicity and the vast availability of specialized libraries. It supports various frameworks like TensorFlow and PyTorch for building complex models. Its readability makes it accessible for researchers and developers alike.

Correct Answer: D. Recommendation Engines

Recommendation engines use machine learning to analyze your past behavior and predict future preferences. They find correlations between your history and that of other similar users. This personalization is a key feature of modern digital platforms.

Correct Answer: A. Cleaning and organizing raw data before training

Data preprocessing ensures that the information fed into an AI model is consistent and error-free. It involves removing duplicates, handling missing values, and scaling data points. High-quality data is the most important factor for a successful AI model.

Correct Answer: C. Create new content like text, images, or music

Generative AI models are trained to produce new data that resembles the training set. Examples include tools that generate realistic art or coherent paragraphs of text. This field has seen rapid growth with the development of Large Language Models.

Correct Answer: D. The difficulty in explaining how an AI reached a specific decision

The black box problem refers to the lack of transparency in how complex neural networks process information. It makes it hard for humans to understand the exact logic behind an AI's output. Explainable AI (XAI) is a field dedicated to solving this specific issue.