构建预测模型，将法语句子分为不同的难度（A1-C2），并构建迷你Web App，根据输入的关键字和难度返回法语Youtube视频的结果。
Built a prediction model to classify the French sentences into different difficulties (A1-C2), and constructed a mini Web-App to return the results of French Youtube videos based on the input keywords and difficulty. (2024-03-23, Jupyter Notebook, 0KB, 下载0次)
使用Jupyter笔记本开发了SMS垃圾邮件分类器，准确率为97.97%。分类器通过从训练数据中学习，将文本消息识别为垃圾邮件或ham并进行分类。
Developed an SMS spam classifier using Jupyter Notebook with an accuracy rate of 97.97%. The classifier identifies and categorizes the text messages as either spam or ham by learning from the training data. (2024-03-08, Jupyter Notebook, 0KB, 下载0次)
LoRa E5 AT命令的除锈器
Rust driver for LoRa E5 AT Commands (2024-03-05, Rust, 0KB, 下载0次)
洛拉E5
LoRaE5 (2024-02-01, C++, 0KB, 下载0次)
使用NLP单词嵌入模型的搜索引擎：SkipGram和GloVe。
Search Engine using NLP word embedding models: SkipGram and GloVe. (2024-01-25, Jupyter Notebook, 0KB, 下载0次)
单词嵌入-属性、意义和训练
Word Embeddings - Properties, Meaning and Training (2024-01-17, Python, 0KB, 下载0次)
为Kaggle源SMS数据设计了NLP管道，通过机器学习实现98%的准确性，通过深度学习实现97%的准确性。实现了高级文本预处理，展示了NLP和机器学习方面的专业知识。通过传统和深度学习方法在提高准确性方面取得了显著的成功。
Designed an NLP pipeline for Kaggle-sourced SMS data, achieving 98% accuracy with machine learning and 97% with deep learning. Implemented advanced text preprocessing, showcasing expertise in NLP and machine learning. Notable success in accuracy improvement through traditional and deep learning methods. (2023-12-30, Jupyter Notebook, 0KB, 下载0次)
开发了一个强化学习人工反馈（RLHF）模型，以解决与印度尼西亚航运电子商务产品相关的挑战，情感分析的验证准确率达到97.79%，航运问题分类的验证准确度达到96.90%，RLHF模型本身的验证正确率达到83.33%。
Developed a Reinforcement Learning Human Feedback (RLHF) model to address the challenges associated with shipping e-commerce products in Indonesia, achieving a validation accuracy of 97.79% for sentiment analysis, 96.90% for shipping problem classification, and 83.33% for the RLHF model itself. (2023-12-28, Jupyter Notebook, 0KB, 下载0次)
r113-tp5-galerie-et-json-a1-Claude2003由GitHub Classroom创建
r113-tp5-galerie-et-json-a1-Claude2003 created by GitHub Classroom (2023-12-05, HTML, 0KB, 下载0次)
学习[https: github.com datawhalechina llm- universe后的课程作业](https: github.com datawhalechina llm- universe%E5%90%8E%E7%9A%84%E8%AF%BE%E7%A8%8B%E4%BD%9C%E4%B8%9A)
Study [https: github.com datawhalechina llm - course assignment after universe] (https: github.com datawhalechina llm - universe% E5% 90% 8E% E7% 9A% 84% E8% AF% BE% E7% A8% 8B% E4% BD% 9C% E4% B8% 9A) (2023-11-29, Python, 0KB, 下载0次)
使用深度学习从英语到法语的翻译模型，准确率为97%。，
A translation model from English to French using deep learning with an accuracy of 97%., (2023-10-25, Others, 0KB, 下载0次)
Yelp数据集上情感分类的非神经模型。第一次作业的第一部分，作为......研究生课程的一部分...，
Non-neural models for sentiment classification on the Yelp dataset. First part of the first assignment as a part of the graduate course on Natural Language Processing (COL772). (2019-02-09, Python, 0KB, 下载0次)
LoRa基板，用于流行的Hoperf RFM95 96 97 98模块，
LoRa Base Board for popular Hoperf s RFM95 96 97 98 modules, (2023-08-28, Others, 0KB, 下载0次)
第772 A1列，
COL772 A1, (2023-08-30, Python, 0KB, 下载0次)
基于STM芯片STM32WLE5JC的SeedStudio模块Grove-Wio-E5的高级应用，该芯片连接到Arduino Nano 33 BLE Sense板...，
Advanced application of SeedStudio module Grove-Wio-E5 based on chip STM32WLE5JC from STM connected to an Arduino Nano 33 BLE Sense board. Works with any Arduino that supports UART (2023-08-24, C++, 0KB, 下载0次)
a1_答案AI，，
a1_answer_AI,, (2023-08-18, Python, 0KB, 下载0次)
锈蚀中的不一致机器人
discord bot in Rust (2023-04-13, Rust, 866KB, 下载0次)
“将线性混合模型的先验转换为重复测量方差分析和配对t检验”的补充材料
Supplementary material for "Translating priors from linear mixed models to repeated-measures ANOVA and paired t-tests" (2022-12-16, HTML, 14562KB, 下载0次)
一种具有改进的池层BF-max-pooling的通用句子编码器
A universal sentence encoder with a improved pooling layer named BF-max pooling (2018-05-18, Python, 274KB, 下载0次)
域视图建模语言（DVML）分配
Domain Views Modeling Language (DVML) Assignment (2015-11-08, Java, 267KB, 下载0次)