Can accurately "guessed" the player psychology, the depth study makes the game more intelligent

"Editor's note" with the development of deep learning technology, the current depth of learning is not only to understand our language and identify our voices, the United States North Carolina State researchers have established a deep learning system, which can be in an open video game environment with 63% accuracy to predict the player's goals, Lenovo to Google's previous acquisition of 400 million U.S. dollars DeepMind, it is not difficult to find in-depth learning in the field of video games is also increasing competition.

The following is the translation:

Deep learning is hot now, though most of the time we know it can identify specific objects in a picture, understand our language or recognize our voices. But a new study by the North Carolina State University suggests that a deep learning model might also make our games smarter.

The researchers used a depth learning model (an accurate denoising technique) to predict a player's goal in a video game that did not set any challenges, specified activities, or fixed paths. As you can imagine, the AI engine in the game will gradually increase the difficulty of the obstacle according to the actual situation, or make proper guidance, and calculate the time when a player actually completes a task.

The researchers used five of the data generated during player games-action status, reasons for action, location, narrative status, and earlier goals-that their systems can predict the goals that these players want to achieve with 62.3% accuracy. It is understood that this technology has improved by nearly 29% compared with the previous accuracy.

It is important that previous technologies require people to manually write the features that their models need to analyze, and that deep learning techniques can learn the characteristics associated with them.

Although this study is not a revolutionary breakthrough for the gaming industry, it may mean that pattern recognition applications based on depth learning are beyond our expectations. In fact, Google's January acquisition of $400 million DeepMind has trained a deep learning system that can learn the rules of video games simply by analyzing the gameplay data.

Original link: New suggests deep learning could improve AI in video games (Compile/Wei revisers/Zhonghao)

Free Subscription "CSDN cloud Computing (left) and csdn large data (right)" micro-letter public number, real-time grasp of first-hand cloud news, to understand the latest big data progress!

CSDN publishes related cloud computing information such as virtualization, Docker, OpenStack, Cloudstack, data center,     sharing Hadoop, Spark, Nosql/newsql, HBase, Impala, Large data viewpoints, such as memory calculation, stream computing, machine learning and intelligent algorithms, provide services such as cloud computing and large data technology, platform, practice and industry information.           &NBSP