Machine Learning is a subfield of Artificial Neural Networks.

• TRUE

• FALSE

Answer explanation

Reference Source: https://www.researchgate.net/publication/354124420_A_Comprehensive_Review_on_Radiomics_and_Deep_Learning_for_Nasopharyngeal_Carcinoma_Imaging#pf5

[QUOTE] The Principle of DEEP LEARNING (DL)

For a better understanding of Deep Learning, it is necessary to clarify the two terms of AI and machine learning, which are often accompanied by and confused with Deep Learning.

The concept of AI was first proposed by John McCarthy, who defined it as the science and engineering of intelligent machines. In 1956, the AI field was first formed in a Dartmouth College seminar.

Currently, the content of AI has become much richer to include knowledge representation, natural language processing, visual perception, automatic reasoning, machine learning, intelligent robots, automatic programming, etc.

The term AI has become an umbrella term. Machine learning is a technology used to realize AI. Its core idea is to use algorithms to parse and learn from data, then make decisions and predictions about events in the real world, which is different from traditional software programs that are hard-coded to solve specific tasks.

The algorithm categories include supervised learning algorithms, such as classification and regression methods, unsupervised learning algorithms, such as cluster analysis and semi-supervised learning algorithms.

Deep Learning is an algorithm tool for machine learning. It is derived from an Artificial Neural Network (ANN), which simulates the mode of human brain processing information, and uses the gradient descent method and back-propagation algorithm to automatically correct its own parameters, making the network fit the data better.

Compared with the traditional Artificial Neural Network (ANN), Deep Learning has more powerful fitting capabilities owing to more neuron levels.

According to different scenarios, Deep Learning includes a variety of neural network models, such as Convolutional Neural Networks (CNNs) with powerful image processing capabilities, Recurrent Neural Networks (RNNs), which primarily process time-series samples, and Deep Belief Networks (DBNs), which can deeply express the training data.

In recent years, CNN-based methods have gained popularity in the medical image analysis domain.

In the studies of NPC imaging using Deep Learning models, CNN was adopted in almost all studies. [/QUOTE]

Which of the following belong to Machine Learning?

• Supervised Learning

• Unsupervised Learning

• Reinforcement Learning

• Knowledge Graph

When shall we use machine learning to solve a given real world problem?

• When the problem changes over time

• When we can't explain human expertise in the form of programming logic

• When humans lack of expertise and need to rely on historical data

• When Google, Microsoft and Amazon are betting on Machine Learning as the next big thing

Deep Learning is a subfield of Machine Learning.

• TRUE

• FALSE

Why did deep learning take off only in the most recent decade?

• Big data - Having more data due to the digitization of data

• The advancement and affordability of digital storage

• The rise of computing power (specialized processors such as graphics processing units or GPUs)

• Because Geoffrey Hinton, Yoshua Bengio and Yann LeCun only started researching into Deep Learning in recent years

Traditional software in the form of Rule-based system uses handcrafted programming logic as knowledge representation.

• TRUE

• FALSE

Unsupervised learning looks for previously undetected patterns in a data set with no pre-existing labels and with a minimum of human supervision.

• TRUE

• FALSE

Which of the following are TRUE about Supervised Learning?

• Supervised learning is the machine learning task of learning a function that maps an input to an output based on example input-output pairs.

• Supervised learning algorithms are trained using labeled data.

• In Supervised Learning, we normally have two sets of problems, Classification and Regression.

• Supervised learning requires minimum human annotated data

Reinforcement Learning models learn by a continuous process of receiving rewards and punishments on every action taken.

• TRUE

• FALSE

As opposed to Supervised Learning, Unsupervised Learning and Reinforcement Learning do not rely on human annotated data (data labeled by human).

• TRUE

• FALSE