Jordan, an authoritative professor of artificial intelligence at the University of California, Berkeley (University of California, Berkeley), published his opinion on this wave of artificial intelligence with the title of "The revolution of artificial intelligence has not yet come". Hot views, in which he mentions that today's public discourse on artificial intelligence often prevents us from seeing a picture of the whole, and its opportunities and risks. When artificial intelligence jumped onto the news page and became the new favorite of technology companies, why should the authority of the academic world jump out to remind us that we still have to work hard on research and development? Perhaps, before discussing the promise of artificial intelligence, we must redefine what this movement seen as an AI renaissance has accomplished.
Deep learning, as everyone knows it, is the hottest topic in this renaissance, but it's really just a small piece of the AI pie. The largest subset of artificial intelligence is machine learning. What is machine learning? Simply put, the most important elements whatsapp list of machine learning are data and algorithms. Data is divided into labeled data (code L) and unlabeled data (unlabeled data, code U). The former contains images and label information (for example, there are cats and people in them), while the latter only has simple images.
Next we need an algorithm (learning algorithm), for example, deep learning is an algorithm. Among them, if only the labeled data (L) is put into the algorithm, it is called supervised learning; on the contrary, if only the unlabeled data (U) is put into the algorithm, it is unsupervised learning (unsupervised learning); both (L + U) are Semi-supervised learning.