Report for: A comparative evaluation of the generalised predictive ability of eight machine learning algorithms across ten clinical metabolomics data sets for binary classification

RT @BroadhurstDavid: This is a big problem in #metabolomics too. I despair at the lack of confidence intervals presented in ML models and i…

05 May 2023

RT @BroadhurstDavid: This is a big problem in #metabolomics too. I despair at the lack of confidence intervals presented in ML models and i…

05 May 2023

This is a big problem in #metabolomics too. I despair at the lack of confidence intervals presented in ML models and in particular PLS-DA models. Go read our paper https://t.co/5O6Qfpxgan

05 May 2023

RT @BroadhurstDavid: Totally 100% agree. As illustrated in our Metabolomics paper https://t.co/5O6QfpOjcn I get depressed every time I see…

22 Jul 2021

RT @BroadhurstDavid: Totally 100% agree. As illustrated in our Metabolomics paper https://t.co/5O6QfpOjcn I get depressed every time I see…

22 Jul 2021

Totally 100% agree. As illustrated in our Metabolomics paper https://t.co/5O6QfpOjcn I get depressed every time I see a ROC curve or coefficients quoted (visualised) without confidence intervals & don’t get me started on people who think doing multi-om

22 Jul 2021

RT @BroadhurstDavid: @Bryanomic @EMN_MetSoc 2/2 for classification ROC curves are a more sensible metric. And remember you need to calculat…

12 Aug 2020

@Bryanomic @EMN_MetSoc 2/2 for classification ROC curves are a more sensible metric. And remember you need to calculate confidence intervals for predations and coefficient values. Read this: https://t.co/5O6QfpOjcn

12 Aug 2020

Great work from @KevinMMendez @StaceyReinke @BroadhurstDavid !

24 Jul 2020

RT @AnupriyaTripat4: A very well-written article comparing the generalizability of binary classifiers in metabolomics data context: https:/…

16 Apr 2020

RT @AnupriyaTripat4: A very well-written article comparing the generalizability of binary classifiers in metabolomics data context: https:/…

16 Apr 2020

RT @AnupriyaTripat4: A very well-written article comparing the generalizability of binary classifiers in metabolomics data context: https:/…

15 Apr 2020

A very well-written article comparing the generalizability of binary classifiers in metabolomics data context: https://t.co/BXUCgXk7MH I got to it a bit late but highly recommend if you haven't read it yet!

15 Apr 2020

RT @StaceyReinke: @polcastellano_ @RoyGoodacre @NutriMetabolom You might also be interested in our work comparing the generalised predictab…

11 Apr 2020

RT @BroadhurstDavid: And if, after that, you are still thirsty for more juicy Data Science fruit, then check out our paper comparing 8 (ye…

RT @BroadhurstDavid: And if, after that, you are still thirsty for more juicy Data Science fruit, then check out our paper comparing 8 (ye…

RT @BroadhurstDavid: And if, after that, you are still thirsty for more juicy Data Science fruit, then check out our paper comparing 8 (ye…

"thirsty"? Greedy 😎

RT @BroadhurstDavid: And if, after that, you are still thirsty for more juicy Data Science fruit, then check out our paper comparing 8 (ye…

And if, after that, you are still thirsty for more juicy Data Science fruit, then check out our paper comparing 8 (yes EIGHT!) #machinelearing algorithms across 10 (yes TEN!) @metabolmics data sets. A total of 80 FREE! Jupyter notebooks. #ShamelessPlug ht

Two interesting read for today in #metabolomics #MachineLearning #NeuralNetworks . 1. https://t.co/DxNWSny4nz 2. https://t.co/O83mHfmdTq #omics #datadriven #bioinformatics #Systems #biology

RT @BroadhurstDavid: @drupad_t @dbkell I respectfully disagree with your conclusion. In human studies N=15 is terrible whatever method you…

Any views on bootstrapping errors? n < 15 will continue in many cases re simple paucity of individuals...

@drupad_t @dbkell I respectfully disagree with your conclusion. In human studies N=15 is terrible whatever method you use. That is more of an issue of sampling error than which #ML you use. The subject of sample size in #ML is addressed in our @metabolomic

@BroadhurstDavid May be, authors can chime in @_e_evans @ejalm @david_sontag . They should also cite your paper https://t.co/AA33YfETpc

04 Feb 2020

RT @StaceyReinke: @polcastellano_ @RoyGoodacre @NutriMetabolom You might also be interested in our work comparing the generalised predictab…

25 Jan 2020

RT @StaceyReinke: @polcastellano_ @RoyGoodacre @NutriMetabolom You might also be interested in our work comparing the generalised predictab…

25 Jan 2020

@polcastellano_ @RoyGoodacre @NutriMetabolom You might also be interested in our work comparing the generalised predictability of 8 ML methods (binary outcome only): https://t.co/tK9vfmSqL4

25 Jan 2020

@Pdorrestein1 @dk_barupal For our recent comparison paper ( https://t.co/oDO0KvFV0l ) we struggled to find 10 publicly available clinical data sets that were sensibly formatted (as an annotated flat text file) with all the required metadata (clinical outco

16 Jan 2020