Artificial Intelligence (A.I.)
Here are a few definitions that define artificial intelligence:
- the capacity of a computer to perform operations analogous to learning and decision making in humans, as by an expert system, a program for CAD or CAM, or a program for the perception and recognition of shapes in computer vision systems.
- the science of making computers do things that human beings can do.
- the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
AI is most likely the broadest concept to describe advanced, computer intelligence. This goes from the use of a computer to imitate the cognitive functions of humans to voice recognition technology such as Alexa or Siri. When machines perform tasks based on “intelligent” algorithms, this is also considered as AI.
“In 1956 at the Dartmouth Artificial Intelligence Conference, the technology was described as such: “Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”|
AI is mainly categorized into two types, type-1 based on capabilities and type-2 based on functionalities.
Type-1: Based on capabilities
- Weak AI or Narrow AI: Weak AI is the most common and available type of AI in the world of Artificial Intelligence. This would be interpreted as performing a task with intelligence. However, this type of AI would not be able to perform any tasks beyond its limitation. It is trained to perform one single specific task. A few examples of weak AI are the self-driving car, image, and language recognition, playing a game such as chess.
- General AI: General AI would entitle being to perform any intellectual task with the same efficiency as a human. In other words, general AI could potentially be smarter and think just like a human. A brain outside a human you could say! This type of AI is what researchers are currently trying to develop.
- Super AI: Super AI would be the type of AI that surpasses human intelligence. Examples of this are the ability to think on its own, resolve, and reason, make judgments, play, learn, communicate on its own.
Type-2: Based on functionalities
- Reactive Machines: Most basic type of AI. These machines only focus on current actions and react with the best possible outcome.
- Limited Memory: Limited memory machines can store past experiences and data for a small period of time. For example, self-driving cars. These cars can store the recent speeds and distance of nearby cars, information to navigate the road, and so on.
- Theory of mind: Theory of mind machines have the ability to understand human emotion, beliefs, and interact with a human.
- Self Awareness: This type of AI would be the future of AI. With a mind of its own; a consciousness, feelings, and self-awareness. These machines will be superior to any kind of human intelligence. This concept is still hypothetical.
Machine Learning (ML)
Here are a few definitions that define machine learning:
- the capacity of a computer to learn from experience, i.e. to modify its processing based on newly acquired information.
- the process of computers changing the way they carry out tasks by learning from new data, without a human being needing to give instructions in the form of a program.
Machine learning is a subfield of AI. The core principal with machine learning is that it “learns” from data on their own. “It’s currently the most promising tool in the AI kit for businesses. ML systems can quickly apply knowledge and training from large data sets to excel at facial recognition, speech recognition, object recognition, translation, and many other tasks.” ML can recognize patterns of its own and make predictions. Resulting in decision making with minimal human supervision. Comparatively to hang-coding everything and calculating all possible outcomes along with instructions on the reaction from the action.
Here are a few definitions that define deep learning:
- a type of artificial intelligence that uses algorithms (= sets of mathematical instructions or rules) based on the way the human brain operates.
- a form of machine learning composed of algorithms that permit software to train itself to perform tasks by exposing multilayered neural networks to vast amounts of data.
- a type of machine learning concerned with artificial neural networks allowing advanced pattern recognition.
Deep learning mimics the human brain by processing data and creating patterns to make decisions. Deep Learning is deeper than ML and is also a subset of machine learning. It “[…]has networks capable of learning unsupervised from data that is unstructured or unlabeled. Also known as deep neural learning or deep neural network.”
How does deep learning work?
“Computer programs that use deep learning go through much the same process. Each algorithm in the hierarchy applies a nonlinear transformation on its input and uses what it learns to create a statistical model as output. Iterations continue until the output has reached an acceptable level of accuracy. The number of processing layers through which data must pass is what inspired the label deep.”
As mentioned above, traditional machine learning, the learning curve is supervised and the programmer needs to be very specific when telling the computer to look for when it finds the image of contains a house or not a house. This process is “feature extraction” and the success rate of this depends solely on the skills of the programmer. Defining the variable “cat”. However, with deep learning, it programs its settings on its own without the help of the programmer. This makes the process more efficient, and surprisingly more accurate.
Interested in knowing more about us and our A.I. tech? You can learn about how it can help grow your App here!
At the start, the program might arrive with training data, and a set of images with meta tags set as “cat” and “not cat”. Further on, the program uses this information to create a set for houses and build a predictive model. With the cat example, the program is not aware of the label “four legs” or “tail” and might predict that anything that has four legs should be a cat. It the program needs to do is look over the pattern pixels in the digital data. With each iteration, each model becomes more and more complex.
As accuracy is an important factor today, immense amounts of training data and processing power is necessary. Since deep learning programming can “[…] create complex statistical models directly from its own iterative output, it can create accurate predictive models from large quantities of unlabeled, unstructured data. This is important as the internet of things (IoT) continues to become more pervasive, because most of the data humans and machines create is unstructured and is not labeled.”
Example of deep learning can include language processing, language translations, medical diagnosis, stock market signals, etc.
- The Anatomy of an Ad Campaign For a Mobile App
- Mobile Fraud Trends at Mapendo in 2019
- How to Successfully Grow Your Mobile App Subscription Business
- The Life of an Account Manager at Mapendo
Visit our site or get in touch with us: