Until now, the most simple task to be done on a computer still requires extremely complex and precise instruction instructions.
Who else around us remember how to program with punch cards? Who else will use DOS?
Computer programming languages ​​have evolved over the years, but the biggest step now is to eliminate complex programming. In other words, the church computer self-study is called machine learning.
Machine learning is a very promising technology. Its ability is a dramatic improvement. In the near future, it will affect each of us and every field in a real and subtle way. Because of this, there are a few things I think everyone should understand.
What is it?
Unlike previous methods that accurately indicate how a computer should be executed to solve a problem, when using machine learning, the programmer does not need to tell 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 those patterns. If you are interested, you can open a link to the website here and have a visual introduction to how machine learning works.
The study of machine learning began as early as the 1950s, when computer scientists came up with the idea of ​​teaching computers to play chess. After that, as computing power increases, computers can identify complex patterns and can therefore make predictions and solve problems.
Machine learning algorithms usually give a set of "teaching" data and then ask for data to answer questions. For example, you might have provided a set of photos for computer instruction, some of which would say "This is a cat" and others would say "This is not a cat." Then you can show the computer a series of new photos, and it will start to judge which ones are cat photos.
Machine learning is constantly increasing its "teaching" data set. Regardless of whether it is right or wrong, each recognized photo will be added to the data set, so the program will become more "smart" and more Complete its mission well.
In fact, this is the learning process.
What is the charm?
Computers can now boldly enter any area that is relevant to us. Although the technology is not perfect in many cases, because of the special concept of machine learning, it can tirelessly improve its performance. In theory, there is no ceiling, it will only get better and better.
As an example of a cat's photo taken before, the computer can now "see" the image and classify it, and can "read" the text and numbers in the picture, even identifying someone or somewhere. Not only do they have the ability to read text, but they can also judge whether the emotions represented are positive or negative by understanding the context.
In addition, the computer can 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 have also learned to write, and machine learning algorithms have been used to write everyday news articles, mainly in areas that require large amounts of data, such as financial and sports reports. This can affect a wide range of tasks, including data entry and classification, that require manual intervention. If a computer can recognize something - such as an image, a document, a file, etc., if the description is accurate, there may be a lot of automation purposes.
| Application status
People can already use machine learning algorithms to achieve a lot of exciting things.
A recent study on the use of computers for assisted diagnosis (CAD) analyzed early scans of women with breast cancer, and the results showed that the computer had 52% of the confirmed time earlier than a year. And, based on the large population, machine learning can learn to understand the virulence factors. Medecision has invented an algorithm that allows it to locate and identify eight signals, allowing diabetics to avoid unnecessary hospitalization.
In addition, presumably you have had such experience, but after an online shop, there is no pickpocket, but in the next few days, the webpage is surrounded by keyword recommendation advertisements that you have searched for. These are just machine learning. The tip of the iceberg of the application. In other cases, such as commercial companies sending coupons to customers, providing product introductions, and recommending new products, they can use the "personalized" super algorithm, all of which has a small purpose, which is to recommend consumers more easily. s product.
Natural Language Processing (NLP) is being used in a variety of interdisciplinary novelty applications. A machine learning algorithm using natural language can replace the customer service specialist and be able to inform the customer of the information they need more quickly. It has also been used to translate obscure wording in contracts into plain language, helping lawyers to organize large amounts of information while preparing a case.
IBM recently surveyed executives at top automakers, and 74% of them expect to see smart cars on the road by 2025.
Smart cars can not only integrate into the entire IoT system, but also learn from its owners and the surrounding environment. It can adjust its internal settings (temperature, music, seat position, etc.) based on driver information, and even automatically fix problems, autopilot, and provide real-time advice based on traffic and road conditions.
| Future development
The imagination space that machine learning brings us is huge, and some of the exciting possibilities include:
Personalized medicine creates unique medical care and treatment plans for users based on their genetic makeup and lifestyle.
Data security, programs can automatically detect malware, viruses, and attacks with high accuracy.
Computer-aided security can predict threats in public places such as airports and stadiums, and check security personnel for missing things.
Self-driving cars can navigate by themselves to avoid traffic accidents.
Advanced fraud detection to protect financial security in the financial and insurance sectors.
Even a “general translation assistant†can translate what you say on your phone or other device in real time, accurately and quickly.
| What is the relationship with me?
For many people, whenever technology advances, they simply welcome new technologies and don't care too much about how they work and the scenarios behind them. But what I want to remind is that we should all care about machine learning because it will bring a lot of benefits to our lives and may change our workforce structure.
Almost everyone on the planet is generating more and more data, and when people use machine learning to do it at work, everything will be subverted. Yes, for many people, these new technologies will make the job easier, but it may also eliminate a lot of work. The algorithm can now help us reply to emails, interpret medical images, find legal cases to win, analyze our data, and more.
Machine learning algorithms rely on "learning" experience from past examples, so that programmers can be saved from endless code without having to worry about unexpected situations. This learning ability, combined with the superiority of robotics and mobile technology, means that computers can now help humans accomplish more complex tasks faster than ever before.
The World Economic Forum estimates that in the next five years we will have 5 million jobs replaced by computers and robots.
This means that no matter what your job is—from lawyers to diagnostics experts, from customer service representatives to truck drivers, you must be aware of how machine learning will affect your field, the business you are in contact with, and what you are doing. jobs. In order to avoid being shocked by the disruptiveness brought by the computer, the best way is to actively understand and prepare from now on.
Stainless Steel Spring Wire,Stainless Steel Welding Wire,Stainless Steel Flux Core Wire,Stainless Steel Woven Wire
ShenZhen Haofa Metal Precision Parts Technology Co., Ltd. , https://www.haofametal.com