In what areas can you program artificial intelligence?

Question

Welcome! I really enjoyed programming artificial intelligence in my research - neural networks, expert machines, and others. But in my work, I mainly develop web applications.

And now I’m thinking about returning to such programming, maybe in a hobby or maybe in a job. Are there areas where AI is commonly used in application development, and a programmer with these skills can look for work?

Or maybe I can sell some ideas to my boss and use AI to expand some of our applications.

What do you experience and ideas using AI in applications?

Answer 1

Recently I started reading the book " Programming the collective mind . " This is a great book that discusses exactly what you are looking for - using AI techniques in web applications.

The book is written clearly, easy to understand, explains everything from the point of view of real applications (it talks about how some commonly used technologies work: Google Pagerank, Amazons recommendation system, match-finding websites, link recommendation systems, Bayesian filters spam and much more), and also uses really useful examples using real data (ebay API, facebook API, etc. are used to collect data). In one chapter, he even explains how to draw graphs optimally (I mean the data structure, not the histograms / line graphs, etc.) (so that the nodes are not too close together, minimal overlapping lines, etc. .), Which can be useful for eg mapping social networks.

I would recommend taking a look at it and see how AI can be applied to web applications.

Answer 2

As a counter-example, analyzing data from water testing equipment is likely to be a bad place to use artificial intelligence:

The Daily WTF: No, we need a neural network

Just a reminder for all of us to choose the right tool for the right job.

Answer 3

You could probably analyze the behavior of visitors to your web applications; how they navigate inside the website to provide a better, optimized interface. Now it depends on which web applications you are running. For online shopping, you can come up with offers extrapolated to customer habits.

You may also find abnormal behavior and fraud. Detecting fraud and bots can use AI.

Answer 4

Neural networks are great for working with images, so one of the areas of web applications for which you could use AI is to identify and / or manipulate image templates on large data sets. For example, a site such as Flickr or Facebook may have interesting educational material for identifying people based on faces or associating pixel groups (those that are functions that you work with) with certain elements mentioned in captions or tags.

There are many things in terms of text manipulation, but usually it is icing on the cake for other web applications. I speak mainly in areas of automatic completion in search bars and background things that the user usually does not see, such as automatic machine translation or improved search capabilities.

The problem with installing AI at the beginning of the application’s proposal is that, as a rule, artificial intelligence is not an end in itself, but rather a way of matching large amounts of data without regular requests from the designer. In general, the user will communicate with the application on a one-to-one basis and, therefore, evaluates it only by the quality of a relatively small number of responses.

Answer 5

Email spam filtering systems - definitely.

Any other security applications that should define patterns for malicious material.

Answer 6

Forecasting , of course.

It is of tremendous value to enterprises (i.e.: inventory optimization) and is especially valuable during the global crisis.

Answer 7

Games need AI.
Expert systems too.

Answer 8

Outside of games, I have seen very little commercial use of AI.

Theoretically, it can be very useful in industrial robotics and visualization, but these areas also tend to be very conservative and inconvenient using non-deterministic algorithms.

You might want to learn what iRobot does, but even they use fairly simple algorithms in their commercial robots.

Answer 9

In the field of cognitive architectures (e.g., Soar, ACT-R, etc.), instead of focusing on algorithms such as A * and games, researchers are exploring patterns of human behavior, including decision-making, cultural exchange, and learning. They often focus on cognitive credibility, i.e. How closely the model tracks what a person does, including time, etc.

These systems are typically strictly based on research with limited commercial applications. So far, anyway. Military applications, especially for training, are quite common.

Answer 10

Risk assessment is a pretty good example for neural networks, mainly because they are pretty good at comparing with the sample. Insurance and credit companies use them to a certain extent to determine the risk to the client.

Answer 11

Image processing for cancer! (In fact, we wrote IEEE documents about this, creating algorithms is much more complicated than coding them, so we write papers about the work of other works)

Answer 12

I have done extensive research on the use of artificial neural networks to classify underwater sound sources. The algorithm seemed to work quite well, especially since I devoted most of the work to find out which combination of the Fourier transform coefficient constituted the best set for classification (with a basic analysis of the components).

Answer 13

Anything (seriously):

http://highlevellogic.blogspot.com/2010/09/high-level-logic-rethinking-software.html

The Open Source Logic (HLL) open source project is dedicated to finding and coding high-level logic in which all other AIs (and, in fact, all programming) are suitable. There are serious specific ideas and code. HLL is already an application platform.

Answer 14

Artificial intelligence refers to the ability of machines to perform psychological functions, such as thinking, perception, learning, overcoming shortcomings and making decisions.

Below are six main areas where we can use AI and machine learning:

1. Health. The use of artificial intelligence in care will help solve problems with high barriers to access medical facilities, especially in rural areas, which suffer from poor property and a limited supply of care professionals. This will be achieved through the implementation of use cases such as a medical specialty driven by artificial intelligence, personal treatment, early detection of potential pandemics and a medical specialty, including imaging.

2. Agriculture: AI promises to revolutionize the food industry and meet the high demand for food (by 2050, global demand should be increased by five hundred additional food products and serve another two billion people by 2050). In addition, it has the potential to address issues such as inadequate forecasting of demand, lack of guaranteed irrigation and excessive / misuse of pesticides and fertilizers. Some use cases include improving crop yields through real-time information, expanded detection of attacks by pursuers, and forecasting crop costs to talk about sowing methods.

3. Education: AI can potentially solve the quality and access issues observed in the education sector. Possible uses include expanding and expanding the learning experience through personalized learning, automating and speeding up administrative tasks, as well as predicting the need for student intervention to reduce dropouts or recommend training.

4. Marketing. Marketing is a way to make your products look appealing to attract more customers. He reviews a lot of recordings to offer screenings and films that you simply would like to support your previous actions and choice of films. As the set of information grows, this technology becomes smarter and smarter every day.

5. Production. According to forecasts, the manufacturing business will become one of the largest beneficiaries of artificial intelligence solutions, so develop the Factory of the Future with the help of universal and changing technical systems to change processes and mechanisms, to repeat unfamiliar or amazing things, creating intelligent ideas. a choice.

Areas of impact include engineering (AI for research and development), supply chain management (demand forecasting), manufacturing (AI can lower prices and increase efficiency), maintenance (predictive maintenance and overblown plus use), quality assurance (e.g., vision systems with machine learning) algorithms for identifying defects and deviations in the characteristics of a product), as well as in-plant deliveries and storage.

6. Energy. Options for potential use in the energy sector include modeling and approval of the energy system to reduce unpredictability and increase efficiency in the alignment and use of energy. In renewable energy systems, AI will change the energy storage through intelligent networks provided by good meters, and further improve the reliability and availability of energy from electrical phenomena. As in the manufacturing sector, AI can be further deployed to prophetically serve the network infrastructure.