{"id":19750183,"url":"https://github.com/wepe/machinelearningeveryday","last_synced_at":"2025-02-28T00:45:21.317Z","repository":{"id":87469990,"uuid":"53014976","full_name":"wepe/MachineLearningEveryday","owner":"wepe","description":"record and share my reading everyday","archived":false,"fork":false,"pushed_at":"2016-04-01T11:03:56.000Z","size":3,"stargazers_count":11,"open_issues_count":0,"forks_count":2,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-01-10T21:26:01.150Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/wepe.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2016-03-03T02:47:19.000Z","updated_at":"2019-09-08T16:08:37.000Z","dependencies_parsed_at":null,"dependency_job_id":"2160bc67-c525-411c-a720-69c69b61fac6","html_url":"https://github.com/wepe/MachineLearningEveryday","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FMachineLearningEveryday","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FMachineLearningEveryday/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FMachineLearningEveryday/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FMachineLearningEveryday/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/wepe","download_url":"https://codeload.github.com/wepe/MachineLearningEveryday/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":241080289,"owners_count":19906258,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-12T02:31:17.517Z","updated_at":"2025-02-28T00:45:21.296Z","avatar_url":"https://github.com/wepe.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"##2016 阅读记录\n\n- 2016-03-05\n\nXgboost中两个我之前没有用过的特性，一个是[用户自定义代价函数](https://github.com/dmlc/xgboost/blob/master/demo/guide-python/custom_objective.py)，另一个是[pred_leaf](https://github.com/dmlc/xgboost/blob/master/demo/guide-python/predict_leaf_indices.py),预测的时候设置pred_leaf为True，将对每个样本返回其在每棵树上的leaf index（一共有num_round棵树），可以当成新的特征来用。\n\n\n- 2016-03-07\n\n阅读了[《A Programmer's Guide to Data Mining》](http://guidetodatamining.com/)这本电子书，内容过于简单，两三个小时读完，干货不多。\n\n- 2016-03-08\n\n[Converting categorical data into numbers with Pandas and Scikit-learn](http://fastml.com/converting-categorical-data-into-numbers-with-pandas-and-scikit-learn/)  这篇文章讲了类别特征的处理，文中提到一点需要引起思考：If you have missing values in a binary feature, there’s an alternative representation:-1 for negatives，0 for missing values，1 for positives。It worked better in case of the Analytics Edge competition: an SVM trained on one-hot encoded data with d indicators scored 0.768 in terms of AUC, while the alternative representation yielded 0.778.\n\n- 2016-03-11\n\n[Imbalanced data – Finding Waldo](http://www.financealleycat.com/?p=69)  这篇文章讲了不平衡数据的处理，都是常见的方法（简单采样，合成采样），但是文章最后讲了一个很有趣的处理方式：如果不平衡数据中某个类别的数据非常少，那么也可以把分类问题当成异常值检测的问题（ [anomaly detection](https://en.wikipedia.org/wiki/Anomaly_detection)），只需要检测出异常值就行了。\n\n- 2016-04-01\n\n看了large-scale svm相关的内容，发现一个不错的工具[EnsembleSVM](https://github.com/claesenm/EnsembleSVM)，在准确率不下降的同时减小计算复杂度，对应论文[EnsembleSVM: A Library for Ensemble Learning Using Support Vector Machines](http://jmlr.org/papers/volume15/claesen14a/claesen14a.pdf)，另外一篇cite比较多的论文[Making large-scale support vector machine learning practical](http://dl.acm.org/citation.cfm?id=299104)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwepe%2Fmachinelearningeveryday","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fwepe%2Fmachinelearningeveryday","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwepe%2Fmachinelearningeveryday/lists"}