TensorFlow is Google's second-generation open source artificial intelligence learning system, a built-in framework learning software library for implementing neural networks. At present, TensorFlow machine learning has become a research hotspot. Starting from the basic machine learning algorithm, this paper analyzes the machine learning algorithm and the TensorFlow framework, and builds the environment under the Linux system to simulate the TensorFlow model of handwritten character recognition to realize the recognition of handwritten characters, thus realizing the learning and application of TensorFlow machine learning framework. .
Machine learning is a multi-disciplinary discipline that enables computer simulation or human learning behavior, and reconstructs its own knowledge structure to improve its performance. At the beginning of 2016, AlphaGo defeated Li Shishi with a big score. The concept of AI has entered people's field of vision since then, and machine learning is the core of AI, which is the fundamental way to make computers intelligent. TensorFlow is Google's second-generation artificial intelligence learning system, an open source deep learning system for making AlphaGo.
1 machine learningA simple example can be used to illustrate the concept of machine learning, using k-nearest neighbor algorithm to improve the pairing effect of dating sites [1]. For example, if you want to meet a friend on the dating site now, and the dating site has two pieces of information for each registered user (the percentage of time spent playing video games and the number of frequent flyers earned each year), you want to know you. Who will be interested in it, then you can use machine learning algorithms to build a simple model. You can enter a machine learning algorithm to create a model by inputting two pieces of information (the percentage of time spent playing video games and the number of frequent flyer miles earned each year) from people who think they are attractive, charismatic, and disliked. As shown in Figure 1. When you want to know if a user is someone who is interested in making friends, input information, computer calculations through this model, can give you a predictive answer, which is a classic supervised learning algorithm.
There are many types of machine learning algorithms, and the supervised learning algorithms described in the above examples are just one of them. If you want to achieve this result in another way, you have a bunch of data as above, but do not classify the data, let the algorithm observe the data according to the way the data is dispersed, and find that the data forms some clusters, as shown in Figure 2. In this way, the data can be automatically classified, which is an unsupervised learning algorithm.
There are many algorithms for machine learning. For example, use learning algorithms to judge how much training information you need. What kind of better approximation function can reflect the relationship between data, so that you can get more accurate judgment with the least training information.
Machine learning is when a machine wants to complete a task, through which it continuously accumulates experience, to gradually complete a task with better and worse errors.
BAM Power Transmission Capacitors
BAM power transmission capacitors
RAM/RFM Induction Heating Capacitors,Water Pump Capacitor,Water Cooled Condense,Tank Capacitor
YANGZHOU POSITIONING TECH CO., LTD. , https://www.cnchipmicro.com