Remembering what you have learned and writing it down is a wonderful way to learn.
In a study published in the January 21 issue of Science, researchers asked 200 college students to spend five minutes reading essays on scientific subjects. After that, you are instructed to read it several times as if you pass the exam, you are instructed to spend 10 minutes to create a "concept map" of the material or to write free-form articles about the passage I will.
A week later, the students tested simple answers about what they remember and asked for logical conclusions from these facts. The students who originally wrote the paper had the best results. Next is Kramer, then Concept Mapper.
Then ask the students to draw a conceptual diagram from their memories, and the author of the article will do the best by breaking the students who made the concept map again.
Although the results of this research should be limited, it suggests that search methods such as essay writing are powerful learning tools. In a progressive course that emphasizes so-called excellent methods such as concept maps, it may be underestimated.
However, concept mapping and stuffing is certainly at least somewhat useful. Students using these methods will report a higher level of understanding than writing their own essays when asked to self assess their learning.
Photo: 1) Renato Ganoza, Flickr. 2) Various kinds of test performance decomposed according to learning method. "Meta-cognitive prediction" means what students think they have learned.
Quote: "Creating more learning using the notion of mapping assistance learning rather than searching practice." Author: Jeffrey D. Karpicke and Janell R. Brant. Science, Volume January 31, 2011, No. 331, No. 331
When suggesting the subject of knowledge, people are always looking for the best way to improve their learning. Some people learn by writing, others use visual references, others try to repeat what they have learned from others to enhance their knowledge. Let's infer these processes to the AI agent. In general, the AI program learning method is basically associated with its underlying algorithm. In modern artificial intelligence scenarios, data scientists spend a lot of time doing regularization and optimization tasks to improve their learning process. But what if the AI program can find the best way to learn a particular knowledge topic? In other words, can the AI program learn how to learn? Well, this is the focus of the field of AI, which is known as meta-learning.
The best artists in the world learn their art by copying the works of the people they admire. You can write clean HTML code using the same strategy (this is also a very creative process). The easiest way to learn how to write beautiful code is easy. Read clean code and copy it
The best way to learn to write is to read it. Do not just read the text. I am interested in reading books. Learn and internalize the intent of everything you are reading. While you are reading you can record notes and pens, you can record the words you need to find a specific word, meaning or thesaurus, or the interesting usage of the whole citations or paragraphs you need to review later. In my opinion, reading (believing that many experienced authors agree) is the most important and effective strategy to learn the best.