Comments (3)
My only comment would be that there is quite some variance in the error values that you get between runs (because the optimization problem is non-convex). So to really determine what works better, you'd have to do 100s of runs with both settings and then compare their means (and variances). Moreover, you'd probably want to repeat the experiment on different datasets and with different dataset sizes.
from bhtsne.
You are right. I forgot to mention that I use externally defined initialization for Y so that error and output is same for each run.
from bhtsne.
Even if the Y is fixed in your experiment: who says the outcome of the experiment wouldn't be different if you had used on another initial Y? You'd have to repeat this experiments many times (for many different datasets) to confirm this is a statistically significant improvement.
from bhtsne.
Related Issues (20)
- Help ME! thanks! HOT 1
- Usage of random generator(s) in the source HOT 2
- How can i visualize the image data like this? HOT 1
- bhtsne.py:135: ComplexWarning: Casting complex values to real discards the imaginary part HOT 1
- Butterfly effect HOT 3
- Can not use the python wrapper in Windows
- transposition based on input method HOT 3
- Why is the exact algorithm 10 times faster? HOT 8
- Dimension problem HOT 3
- Can't compile the .exe with visual studio 9.0 HOT 9
- Pytorch version? HOT 4
- python wrapper - Cost for each sample
- Performance difference to the old version HOT 1
- C API HOT 3
- Bhtsne for large datasets HOT 1
- Performance difference Windows/Ubuntu HOT 2
- t-SNE for Java/Scala/Kotlin/Clojure
- Is there a rule of thumb for the lower bound on the perplexity?
- Why use zeroMean in gradient update?
- Finding P and Q matrices
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from bhtsne.