This is seriously my favorite TH-cam channel for Julia programming. I'm definitely going to mention this new playlist on this weeks episode of TalkJulia. Please keep it up! It's wild how much I have learned from watching these well thought out and descriptive videos made by the dabbling doggo.
As clear as usual! It would be very helpful if you can do some videos about using PyCall, Conda.jl, and PythonCall.jl, CondaPkg.jl. In practical application we need to use some python packages or scripts from papers. For example, alphafold, pytorch geometric, rdkit etc. Thank you!
Hey! Is there any benefit to weighting the votes by the distances? Conceptually, a point farther away from the test point should be 'less related' to the test point.
Yes, there are a couple of methods to determine the Class. One method is to use the "majority voting" method which is shown in this tutorial. Another common method is to use "distance weighting" which assigns a higher weight to shorter distances. I haven't actually tried it with the NearestNeighbors.jl package so I can't vouch for how well it works. The Distances.jl package also contains a list of various "weighted" distance metrics that it supports, including WeightedEuclidean and WeightedSqEuclidean, among others.
In VS Code, go to Settings > Julia: Execution: Result Type. In the drop-down menu, select either "inline" or "both". After you do that, place your cursor at the end of a line of code and hit Shift+Enter.
![[05x08] Intro to Artificial Neural Networks with Flux.jl (1 of 2); Julia Supervised Machine Learning](http://i.ytimg.com/vi/zmlulaxatRs/mqdefault.jpg)
![[05x08] Intro to Artificial Neural Networks with Flux.jl (1 of 2); Julia Supervised Machine Learning](/img/tr.png)
Fantastic video and channel. Thanks.
This is seriously my favorite TH-cam channel for Julia programming. I'm definitely going to mention this new playlist on this weeks episode of TalkJulia. Please keep it up! It's wild how much I have learned from watching these well thought out and descriptive videos made by the dabbling doggo.
Glad you're enjoying the series and thank you for your support!
Excelente!!!, saludos desde México, sigue así por favor
¡Saludos! ¡Gracias por ver!
Here I am once again to thank you for sharing your knowledge.
I tested k-NN for regression and the results also were good.
Thank you!
Cool! Good to hear you're getting good results!
very good tutorial
Thanks for watching!
As clear as usual!
It would be very helpful if you can do some videos about using PyCall, Conda.jl, and PythonCall.jl, CondaPkg.jl. In practical application we need to use some python packages or scripts from papers. For example, alphafold, pytorch geometric, rdkit etc. Thank you!
Hey! Is there any benefit to weighting the votes by the distances? Conceptually, a point farther away from the test point should be 'less related' to the test point.
Yes, there are a couple of methods to determine the Class. One method is to use the "majority voting" method which is shown in this tutorial. Another common method is to use "distance weighting" which assigns a higher weight to shorter distances. I haven't actually tried it with the NearestNeighbors.jl package so I can't vouch for how well it works. The Distances.jl package also contains a list of various "weighted" distance metrics that it supports, including WeightedEuclidean and WeightedSqEuclidean, among others.
How did you get that type linting after the variables or expressions you have declared? Such as "100x2 Matrix{Float64}". Cheers
In VS Code, go to Settings > Julia: Execution: Result Type. In the drop-down menu, select either "inline" or "both". After you do that, place your cursor at the end of a line of code and hit Shift+Enter.