Log In Sign Up

Semi-Supervised Learning

What is Semi-Supervised Learning?

Semi-supervised learning is a machine learning technique that labels some of the data in an AI’s database but not all. With this reference benchmark, the technique can infer or learn what the unlabeled data represents with far better accuracy than in unsupervised learning (where no data is labeled), but without the time and costs needed for supervised learning (where all data is labeled).

Why is this Useful?

Semi-supervised learning allows neural networks to mimic human inductive logic and sort unknown information fast and accurately without human intervention. Any problem where you have a large amount of input data but only a few reference points available is a good candidate semi-supervised learning. A classic example is a photo archive with millions of random images. Instead of manually labeling each picture, a human searching for images of people can just tag a few relevant samples from the database. Then the neural network can scour the databank and find every image it believes represents a human.

Practical Uses of Semi-Supervised Learning

  • Cheaper, Faster AI Training

    – By cutting the costs and time needed to train new deep learning networks, you can deploy new programs with reasonable accuracy in a fraction of the time as other learning methods.

  • Computer Vision – Training a deep network to recognize specific reference features gives the machines  image recognition abilities.
  • Enhanced Neural Network Development – By combining semi-supervised learning with other techniques, such as reinforcement, allows data scientists to create incredibly human-like and complex neural networks.