olszewski-otmianowski-how-to-efficiently-model-learners-knowledge-with-pydata-warsaw-2019.json

{
  "description": "During our presentation we will share the results and experiences\nconnected with implementing state-of-the-art techniques for modelling\nlearners knowledge using Recurrent Neural Networks (Deep Knowledge\nTracing).\n\nKnowledge Tracing (KT) is one of the most important research areas in\npersonalized education nowadays. It allows us to trace learners\u2019\nknowledge over time so that we can accurately predict how they will\nperform in the future. By improving the quality of such models we can\nbetter adjust the adaptive learning experience to the needs of\nparticular students. In recent years the idea of using recurrent neural\nnetworks for learners knowledge tracing (Deep Knowledge Tracing, DKT)\ngained a lot of attention, as it has been shown that it generally\noutperforms traditional methods. During our presentations we will share\nthe results and experiences connected with implementing this method in\none of the Pearson personalized learning products. We will focus on\nchallenges that we have encountered during the model development process\nrelated to the framework we\u2019ve used (TensorFlow), training performance,\nexperiment tracking and having multiple people working simultaneously on\nthe same model. We\u2019ll also share the results and compare them with the\nstate of the art results from other papers.\n",
  "duration": 1875,
  "language": "eng",
  "published_at": "2020-01-03T08:00:08.000Z",
  "recorded": "2019-12-13",
  "speakers": [
    "Mateusz Otmianowski",
    "Mikolaj Olszewski"
  ],
  "thumbnail_url": "https://i.ytimg.com/vi/CzRmRZNpB1Y/hqdefault.jpg",
  "title": "How to efficiently model learner\u2019s knowledge with recurrent neural networks",
  "videos": [
    {
      "type": "youtube",
      "url": "https://www.youtube.com/watch?v=CzRmRZNpB1Y"
    }
  ]
}