Explore tweets tagged as #elasticNet
Progress update on @JungleRockRes’ "Alpha on Trend-Following Beta" white paper:. I’ve gotten to a satisfactory result on the ElasticNet Component. But again, not a 1 to 1 mirror of the paper, nonetheless happy overall with the results:
2
2
11
Day 18 of COESIS Knowledge Quest.Today I worked on Black Friday data, implemented feature engineering, and applied regression models like ElasticNet, Lasso, and Gradient Boosting. I built a Web App for sales prediction. #30daysoflearning #COESIS #COESISKnowledgeQuest @coesisnp
0
0
2
New nnetsauce version with CustomBackPropRegressor (CustomRegressor with Backpropagation) and ElasticNet2Regressor (Ridge2 with ElasticNet regularization). - `CustomBackPropRegressor` is based on `CustomRegressor` and allows for a backpropagation of the hidden layer parameters.
0
1
0
👩💻Completed Week 1 Day 5 of ML on @OpenLearn_NITJ cohort 1.0, learned about regularization and why it is important, Lasso, Ridge and ElasticNet for avoiding overfitting (and underfitting ;) ). here are the visualizations from DIY task .great blog by @Ratinder_999 sir!
1
1
9
ちなみにsklearnのElasticNetクラスのalphaとRidgeクラスのalphaとの間には少し違いがあって、Ridgeクラスの損失は.||y-Xw||^2 + α||w||^2.なのに対し、ElasticNetの`l1_ratio=0`における損失は.1/(2n)*||y-Xw||^2 + (1/2)*α'||w||^2.なので、α'=α/nに設定しておかないと両者の解は一致しません。
0
5
29
🤖 Viete, ako používať Lasso, Ridge a ElasticNet regresie?.Čo sa dnes naučíme?.🔹 Aké existujú pokročilejšie typy regresií?.🔹 A ešte omnoho viac z oblasti Machine Learning (strojové učenie) ➡ #vita #itacademy
0
0
1
Day 17 of ML (Yes I managed not to break the learning streak).Learnt-.1) Lasso regression: Reason of sparsity.2) Cost function optimisation comparison.3) ElasticNet regression & it’s performance plots.4) Hyper-parameter importance in ElasticNet regression. Submitted office report
4
4
151
Adriano Zaghi, ML for predicting AMR (antimocrobial resistance): classification and regression, with elasticnet, elasticnet, adaboost. The Bologna cradle of #Bioinformatics rocks!.#BITS2023 #Bari
1
0
10
Day 28 of #100DayofCode.↳ DSA Progress .733. Flood Fill .994. Rotting Oranges .3. Number of Islands .↳ ElasticNet Regression with code .↳ learned about ownership in move .#dsa #java #LearnInPublic #BuildInPublic #AtoZDSA #ML
0
0
5
Day 39 of #100daysofML #100DaysOfCode.1.Learnt about the elasticnet regression.2.Done with views and keys in dbms.3.Maximum width of binary tree.4.Working on my project(scrapped data). Excited to sail through this journey.💪 #100daysofcode
0
0
7
Week 4 Bootcamp Recap (3–9 July):. 📚 Topics Covered:.→ ML Intro & Types.→ Linear Regression .→ Over/Underfitting, OLS, Pipelines.→ Ridge, Lasso, ElasticNet.→ Hyperparameter Tuning.→ Logistic Regression & Metrics.→ End-to-End Project.→ Kaggle Notebook 90% accuracy
Week 3 Recap (26 June–2 July):.→ Probability Distributions.→ Hypothesis Testing, P-value, T/Z tests, Bayes’ Theorem.→ CLT, C.I. , Chi-Square, ANOVA.→ Feature Engineering.→ EDA. 🎯 30% done with @Krishnaik06 sir's bootcamp. #buildinpublic #LearnToCode
2
0
2
A Reminder of an old Modeling Trick(Choosing the model) in #MachineLarning that never dies 🧑💻🤖🚀: . - Linear regression (l1/l2 as norms and elasticnet as regularization) as a starter. - LightGBM for pretty much everything as it’s fast and accurate. - CatBoost if you have
2
0
6
Day 25 Bootcamp Recap (6 July):. 📚 Topics Covered:.→ Ridge Regression.→ Lasso & ElasticNet Regression.→ Types of Cross-Validation.→ Practical implementation & hands-on. #100DaysOfCode #LearnToCode #learninpublic
Day 24 Bootcamp Recap (5 July):.📚 Topics Covered:.→ Overfitting & Underfitting.→ Linear Regression with OLS.→ Polynomial Regression.→ Pipeline in Polynomial Regression. #100DaysOfCode #buildinginpublic #learntocode #LearnInPublic
0
0
0
dev log 5th June.4-5 hours ml. > onsite intern (10 am - 1 pm).> mathematics behind ElasticNet Regression.> did a fun test to compare r2 scores of different regression algorithms on same dataset.> mathematics behind logistic regression
5
2
30
0
0
0
Took some time to get through, but here's. Day-5 (W-1) @OpenLearn_NITJ .Understanding regularization on different cases (underfitting/overfitting models), loss function, Lasso, Ridge and ElasticNet along with balancing Bias and Variance. Was good fun implementing on kaggle.
0
1
8
0
0
0
そういえば、sklearnのLassoやElasticNetは、誤差項をデータ数Nで割ってから正則化項を足すけど、RidgeはNで割らないんですよね。なので、同じalphaの値でも、正則化のインパクトがてんで違う。非本質的な話ではあるけど、結構罠だと思う。（多分、教科書的な記述と整合性を取るためなんだろうけど。）.
0
3
80