Until now, to complete one of the simplest tasks on a computer still requires extremely complex and precise instructions.
Who else around us remembers how to program with punched cards? Who else will use DOS?
Computer programming languages ​​have evolved over the years, but the biggest step that needs to be crossed now is eliminating complex programming. In other words, teaching the church computer to self-learning is called machine learning.
Machine learning is a very promising technology. Its ability is a leap forward. In the near future, it will affect all of us and every field in a real and subtle manner. For this reason, there are several things I think everyone should understand.
What is it
Unlike the previous accurate instructions on how the computer should perform to solve the problem, when using machine learning, the programmer does not need to tell it how it should learn to solve the problem.
Machine learning is essentially a very advanced statistical application that learns how to identify data patterns and make predictions based on these patterns. If you are interested, you can open the website link here and have a visual introduction to how machine learning works.
Machine learning research began as early as the 1950s, when computer scientists came up with ideas on how to teach computer chess. After that, as computing power increases, computers can recognize complex patterns and can therefore make predictions and solve problems.
Machine learning algorithms usually give a set of "teaching" data, and then require the data to answer questions. For example, you may have provided a set of photos for computer teaching, some of which would say "This is a cat" and others say "This is not a cat." Then you can show the computer a series of new photos, and it will start to figure out for yourself what the cats are.
Machine learning is constantly increasing its "teaching" data set. Whether right or wrong, each identified photo will be added to the data set, so that the program will become more and more "smart" and more capable. Good completion of its mission.
In fact, this is the learning process.
| What is the charm
Computers can now boldly enter into any area that is closely related to us. Although the technology is incomplete in many cases, because of the special concept of machine learning, you can tirelessly and constantly improve its performance. In theory, there is no ceiling, it will only become better and better.
As in the example of a photo of a cat that we previously mentioned, computers can now "see" pictures and categorize them. They can also "read" the words and figures in the pictures, and even identify individuals or places. They not only have the ability to read texts, but also can understand whether the emotions represented are positive or negative through understanding the context.
In addition, computers can also listen, understand and respond to us. The virtual assistant in your pocket may be Siri, Cortana, or Google Assistant. This represents a major leap in the ability of computers to understand human natural language and is constantly improving.
Computers now also learn to write, and machine learning algorithms have been used to write some daily news articles, mainly in fields that require large amounts of data, such as financial and sports reports. This will affect a wide range of tasks that require manual intervention such as data entry and classification. If a computer can recognize something - such as an image, a document, a file, etc., if the description is accurate, there may be many automated uses.
| Application Status
People can already use machine learning algorithms to achieve many exciting things.
A recent study on the use of computer-assisted diagnostics (CAD) has analyzed the early scan results of women with breast cancer. The results show that computers have 52% of them diagnosed earlier by about a year. And, based on a large population, machine learning can learn to understand the causative factors. Medecision invented an algorithm that allowed it to locate and identify 8 signals that would allow diabetic patients to avoid unnecessary hospitalization.
In addition, presumably you have had this experience, but after a few visits to the online store, they did not pick up their hands. However, in the days that followed, the keyword-recommended advertisements that you have searched all the way around the webpage are just machine learning. Applied tip of the iceberg. In other cases, such as when a commercial company sends coupons, provides product introductions, or recommends new products to customers, it can play a "personalized" super algorithm. All this has only one small objective, that is, it is recommended that consumers are more easily favored. s product.
Natural language processing (NLP) is being used in various cross-disciplinary novel applications. Machine learning algorithms using natural language can replace customer service specialists and inform customers of the information they need more quickly. It has also been used to translate obscure language in contracts into plain language, helping lawyers organize large amounts of information in the preparation of cases.
IBM recently conducted surveys among executives of top automakers, and 74% of them expect to see smart cars on the road before 2025.
Smart cars can not only integrate into the entire Internet of Things system, but also learn about its owner and its surroundings. It can adjust its own internal settings (temperature, music, seat position, etc.) based on driver information, and can even fix problems automatically. It can also drive automatically, and it can provide real-time advice based on traffic and road conditions.
Future development
The imagination space that machine learning brings to us is enormous. Some of the exciting possibilities include:
Personalized medicine creates unique medical care and treatment plans for users based on genetic makeup and lifestyle.
With data security, the program can automatically detect malware, viruses, and attacks with a high degree of accuracy.
With computer-assisted security, threatening personnel can be predicted in public places such as airports and stadiums, and security personnel can be checked to miss things.
Self-driving cars can navigate themselves and avoid traffic accidents.
Advanced fraud detection protects the safety of funds in the financial and insurance sectors.
Even a "universal translation assistant" can translate what you say to your cell phone or other device in real time, accurately and quickly.
| What does it have to do with me?
For many people, whenever technology advances, they only welcome new technologies and do not care too much about its working principles and behind-the-scenes usage scenarios. But I want to remind that we should all care about machine learning because it will bring many benefits to our lives and it may also change our labor structure.
Everyone on earth is producing more and more data. When people use machine learning to deal with work, everything will be overturned. Yes, for many people, these new technologies will make the job easier, but they may also eliminate a lot of work. The algorithm can now help us reply to e-mails, interpret medical images, find successful legal cases, analyze our data, and more.
Machine learning algorithms rely on "learning" experiences from past examples, allowing programmers to save from endless code without having to think about unexpected situations. This learning ability, coupled with the superiority of robotics and mobile technology, means that computers can now help humans accomplish more complex tasks faster and better than ever before.
The World Economic Forum estimates that 5 million jobs will be replaced by computers and robots in the next five years.
This means that no matter what your job is—from lawyers to diagnostics specialists, from customer service representatives to truck drivers, you must pay attention to how machine learning will affect your field, the business you are exposed to, and what you are doing. jobs. To avoid being shocked by the subversive nature of computers, the best way is to actively understand and prepare for the present.
Optical Rotary Sensor,Custom Encoder,Optical Encoder 6Mm Shaft,Handwheel Pulse Generator
Jilin Lander Intelligent Technology Co., Ltd , https://www.jllandertech.com