Deep Learning
Deep learning is probably one of the hottest tech topics right now. Deep learning is a sub-branch of ML in that it also has a set of learning algorithms that can train on and learn from data, and more specifically Deep learning is powered by neural networks. Moreover, Deep learning can perform outside the machine learning area and comes to assist other areas like computer vision and NLP
Machine Learning(ML) is about training the learning algorithms like Linear Regression, KNN, K-Means, Decision Trees, Random Forest, and SVM with datasets, so that the algorithms could learn to adapt to a new situation and find patterns that might be interesting and important. For training ML, the dataset can be labeled, e.g. it comes with an “answer sheet”, telling the computer what the right answer is, like which emails are spams and which are not. This is called supervised learning and algorithms like Linear Regression and KNN are used for such supervised regression or classification. Other datasets might not be labeled, and you are literally telling the algorithm such as K-Means to associate or cluster patterns that it finds without any answer sheet. This is called unsupervised learning. Turing Test: The computer passes the Turing Test if a human, after posing some written questions to the computer, cannot tell whether the written responses come from another human or the computer. Artificial Intelligence: A Modern Approach, Peter Norvig and Stuart Russell define the 4 capabilities a computer must command in order to pass the Turing Test:
Natural Language Processing: to communicate successfully in English
Knowledge Representation: to store what the computer reads
Automated Reasoning: to use the stored knowledge to answer questions and draw new conclusions
Machine Learning: to adapt to new circumstances and to identify new patterns.
Deep learning requires large amounts of labeled data. For example, driverless car development requires millions of images and thousands of hours of video.
Deep learning requires substantial computing power. High-performance GPUs have a parallel architecture that is efficient for deep learning. When combined with clusters or cloud computing, this enables development teams to reduce training time for a deep learning network from weeks to hours or less.
Examples of Deep Learning at Work.
Deep learning applications are used in industries from automated driving to medical devices.
Automated Driving: Automotive researchers are using deep learning to automatically detect objects such as stop signs and traffic lights. In addition, deep learning is used to detect pedestrians, which helps decrease accidents.
Aerospace and Defense: Deep learning is used to identify objects from satellites that locate areas of interest and identify safe or unsafe zones for troops.
Medical Research: Cancer researchers are using deep learning to automatically detect cancer cells. Teams at UCLA built an advanced microscope that yields a high-dimensional data set used to train a deep learning application to accurately identify cancer cells.
