At the beginning of the course, let’s talk about why a neural network is needed?
Let’s start with the issue of grabbing tickets for the New Year. The time goes back to a few years ago. When the 12306 website was first launched, I believe everyone had talked about tickets on the website. If you grab a ticket on the website, you need to enter a verification code every time you grab a ticket. The user is very annoying, because some verification codes are very strange humans to identify these verification codes and they often make mistakes. If they are wrong, they have to re-enter them again. Obviously, the ticket is seen, but the ticket is stolen by someone else. , How upset is in my heart, this psychological shadow area, everyone think about it, it is just seeing such a user’s pain point, the boss found the product Huang and said an idea, this idea is that the software automatically recognizes the verification code and enters it. There is no need for human identification and input, so that the user can place an order in advance. If the booking is successful, the order will be automatically issued. Wouldn't the user die, the user can let others use it first, and after the dependence is formed, we can Introduce corresponding members and charged products, for example, only after becoming a paying member can you use the function of automatic order placement.
Product Yellow very much agrees with the boss’s idea. Product Yellow also thinks it’s easy to realize this verification code recognition, because the verification code on the 12306 website was only a number, it only recognized the 10 Arabic numerals from 0 to 9, and the product was yellow. think that such a small case on the project do not need to go directly to find developers to get on the line, product yellow recruited development ZHANG, roughly explained under demand, few gave the following code to make pictures ZHANG reference.
Application Zhang received a product yellow after the submission of this demand, wondering for a long time to find the product Huang , said that this matter really can not do it, the product Huang full of doubt looked at the application Zhang said, why you? You give me a reason.
According to Zhang , I have no way to find out all the identification rules of numbers. Take the verification code in the above figure 9530 as an example. The digital image of 9 has some red dots. After these red dots, I must write a program. List the locations of these red dots, otherwise the program will not be able to determine that this is a number 9, but these red dots are randomly generated by the 12306 website. It is impossible for me to write a program to give these unknown conditions, other things There is a similar situation with numbers, so I can't do it at all. After listening to the application of Zhang ’s explanation, the product Huang also found it reasonable, but then I thought about it, it’s not right. It is said that these software products have been launched abroad . After talking about this situation and the application , Zhang said that he also knows this. It’s said that this product uses neural network-based deep learning OCR recognition technology to do the verification code recognition. Zhang suggested that the product Huang finds the algorithm and Liu chats, because it seems that he is studying this neural network and machine. Learn. After Product Huang received Application Zhang ’s suggestion, he went to Algorithm Liu . After listening to the requirements explained by Product Huang, Algorithm Liu said this thing slowly. Both machine learning and deep learning can be done. They are all very good at Dealing with this kind of data seems to be regular to humans and there are a lot of this kind of data to prove the existence of this kind of law, but traditional programming is something that can't be achieved . How did you do it? He also needs further study. He suggests that Product Huang join in while playing AI while gnawing on mathematics.Learn this course. This course is said to allow products, application development, algorithm design, and project management to learn the knowledge that their position needs in the entire development process of AI products.
After listening to Liu 's suggestion from Algorithm Liu , Product Huang immediately signed up for the course "While Playing AI and Chewing on Mathematics" , and also brought in Zhang Tongxie , the application manager of the Development Department, and Li Tongxie, the project manager, to sign up for learning together .
Speaking of this, I will leave a question for everyone: What problems are deep learning and machine learning good at solving, and what are the prerequisites?
In the end, it is still an old routine. If you feel that the course GET for children's shoes has reached you, please click three times!