从这里开始
从这里开始
五 成果导向课程
每门课程都包含明确的成果、能力规划、实践活动和应用毕业设计。
主题曲目
按曲目浏览
跳入您关心的领域。每条赛道都有多个简单的英文指南。
- 91 指南基础知识核心概念：人工智能如何在高层次上工作。探索
- 177 指南语言人工智能LLM、NLP、ChatGPT、提示和文本理解。探索
- 170 指南视觉人工智能图像生成、计算机视觉、视频和设计。探索
- 124 指南音频人工智能语音、音乐、播客和音频优先的人工智能。探索
- 123 指南公司OpenAI、Anthropic、Google、Meta 等的配置文件。探索
- 346 指南应用领域人工智能如何出现在编码、搜索、代理和日常生活中。探索
- 165 指南行业医疗保健、金融、教育、能源等领域的人工智能。探索
- 319 指南技术的芯片、微调、RAG、向量和人工智能基础设施。探索
- 142 指南社会偏见、隐私、版权、安全和人工智能的未来。探索
全库
所有指南
1657 的 1657 显示指南。按曲目过滤或在上方搜索。
- 技术的音频数据增强Audio data augmentation creates label-preserving variations of training recordings to help models handle expected changes in noise, rooms, speed, or encoding.3 最小阅读量阅读
- 技术的超参数的贝叶斯优化Bayesian optimization searches for strong hyperparameter settings by fitting a probabilistic surrogate to completed trials and using an acquisition rule…3 最小阅读量阅读
- 技术的动态时间扭曲Dynamic Time Warping (DTW) measures sequence dissimilarity by finding a low-cost alignment path that can stretch or compress time.3 最小阅读量阅读
- 技术的因式分解机Factorization machines model pairwise feature interactions by representing each feature with a learned low-dimensional factor vector.3 最小阅读量阅读
- 技术的GARCH 波动率模型GARCH models forecast time-varying conditional variance by relating current volatility to past squared shocks and past variance estimates.3 最小阅读量阅读
- 技术的Griffin-Lim Spectrogram InversionThe Griffin-Lim algorithm estimates a waveform from a magnitude spectrogram by iteratively updating phase to make the waveform's short-time Fourier transform…3 最小阅读量阅读
- 技术的隐式反馈和交替最小二乘法隐式反馈推荐器从点击、播放和购买等没有提供明确评级的行为中学习。3 最小阅读量阅读
- 技术的时间序列预测的滞后特征Lag features turn forecasting into supervised learning by pairing past observations and other known-at-forecast-time inputs with a future target.3 最小阅读量阅读
- 技术的NDCG 和排名指标Ranking metrics evaluate the order of items rather than only whether individual labels are correct.3 最小阅读量阅读
- 技术的使用 Piper 和 Kokoro 进行离线文本转语音离线文本转语音使用下载的软件和模型或语音文件在本地将书面文本转换为语音。3 最小阅读量阅读
- 技术的用于源分离的时频掩蔽Time-frequency masking separates an audio mixture by estimating weights over its spectrogram and applying them to the mixture.3 最小阅读量阅读
- 技术的两塔推荐模型A two-tower recommender encodes a user or query and each candidate item separately into vectors, then scores a pair by vector similarity.3 最小阅读量阅读