TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

How to visualize decision trees

343 点作者 parrt超过 6 年前

8 条评论

parrt超过 6 年前
Decision trees are the fundamental building block of gradient boosting machines and Random Forests™, probably the two most popular machine learning models for structured data. Visualizing decision trees is a tremendous aid when learning how these models work and when interpreting models. Unfortunately, current visualization packages are rudimentary and not immediately helpful to the novice. So, we've created a general package called animl for scikit-learn decision tree visualization and model interpretation.
评论 #18080206 未加载
评论 #18080157 未加载
oscilloscope超过 6 年前
There&#x27;s a nice interactive version of a decision tree diagram here, in the section &quot;Growing a Tree&quot;.<p><a href="http:&#x2F;&#x2F;www.r2d3.us&#x2F;visual-intro-to-machine-learning-part-1&#x2F;" rel="nofollow">http:&#x2F;&#x2F;www.r2d3.us&#x2F;visual-intro-to-machine-learning-part-1&#x2F;</a>
评论 #18080778 未加载
评论 #18080327 未加载
benmccann超过 6 年前
I have a model trained with XGBoost Java. Can I take the model file and read it with scikit-learn to visualize it with this library?
jdonaldson超过 6 年前
Good to see others looking into tree model viz. I&#x27;ve done work with larger scale tree visualizations and found you quickly run out of space. I wound up using interactivity to reveal branch level info, dynamically pruned the tree based on train support, and I used a more sophisticated layout technique to pack more info in. <a href="https:&#x2F;&#x2F;www.google.com&#x2F;amp&#x2F;s&#x2F;blog.bigml.com&#x2F;2012&#x2F;01&#x2F;23&#x2F;beautiful-decisions-inside-bigmls-decision-trees&#x2F;amp&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.google.com&#x2F;amp&#x2F;s&#x2F;blog.bigml.com&#x2F;2012&#x2F;01&#x2F;23&#x2F;beaut...</a><p>FWIW visualizing trees like that helps spot problems really quickly. Overfitting behavior typically involves overusing a certain field, or growing long and relatively narrow branches.
评论 #18085452 未加载
parrt超过 6 年前
Just a note that dtreeviz works cross-platform now! Mac, Windows, Linux. &quot;pip install -U dtreeviz&quot; See more at <a href="https:&#x2F;&#x2F;github.com&#x2F;parrt&#x2F;dtreeviz" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;parrt&#x2F;dtreeviz</a>
beamatronic超过 6 年前
If you model the decision tree as a directed graph, there’s no reason you couldn’t export it into a Doom&#x2F;Quake level or even Minecraft or Roblox
评论 #18081427 未加载
nestorD超过 6 年前
Not sure about the choice of pie chart as the default leaf format (humans are bad at guessing proportions from pie charts) but otherwise it does look great and convey the information efficiently.
评论 #18085487 未加载
swinghu超过 6 年前
Good,very informtion distilled from the drawing