Discover and read the best of Twitter Threads about #featureEngineering
Most recents (4)
12 Best FREE #SQL Courses and Certifications Online in 2023 — compiled by @tut_ml
—————
#Databases #DataScience #BigData #Analytics #DataScientists #100DaysOfCode
mltut.com/best-free-sql-…
—————
#Databases #DataScience #BigData #Analytics #DataScientists #100DaysOfCode
mltut.com/best-free-sql-…
#SQL Notes for Professionals: books.goalkicker.com/SQLBook/
[free 166-page PDF download]
———
#Database #DataScience #DataAnalytics #DataScientists #DataProfiling
[free 166-page PDF download]
———
#Database #DataScience #DataAnalytics #DataScientists #DataProfiling
#SupervisedLearning is a type of #MachineLearning where an algorithm is trained on a labeled dataset to predict outcomes for new data
The labeled dataset contains input variables and the desired output, and the algorithm uses this information to make predictions
The labeled dataset contains input variables and the desired output, and the algorithm uses this information to make predictions
In #SupervisedLearning, the algorithm is constantly adjusting its parameters to minimize the prediction error
One of the most popular algorithms for #SupervisedLearning is #LinearRegression, used for prediction problems where the target is continuous
One of the most popular algorithms for #SupervisedLearning is #LinearRegression, used for prediction problems where the target is continuous
Another common algorithm is #LogisticRegression, used for classification problems where the target is binary
#DecisionTrees and #RandomForests are commonly used for both regression and classification problems
#DecisionTrees and #RandomForests are commonly used for both regression and classification problems
#TransmogrifAI #Scala #DSL over #ApacheSpark with rich functionality for #featureEngineering & #featureEvaluation and expressive typelevel / #functionalProgramming & design abstractions
docs.transmogrif.ai/en/stable/deve…
#MachineLearningEngineering
#WorkflowManagement #StateManagement
docs.transmogrif.ai/en/stable/deve…
#MachineLearningEngineering
#WorkflowManagement #StateManagement
Hypothesis: #TransmogrifAI with #workflowManagement done on #Cloudflow (#MLOps over #Kubernetes) could be functionally superior to "#Kubeflow Pipelines" because would allow more complex orchestrations, beyond Spark only, through #Akka
@tovbinm thoughts?↕️
@tovbinm thoughts?↕️
@threadreaderapp unroll
(Thread) Here is a short summary of "Rules of ML" by Martin Zinkevich (developers.google.com/machine-learni…). The #MachineLearning life cycle should consist of three phases:
Phase 0
- Launch the product without #MachineLearning . Collect #Data
Phase 1
- Building the first basic/transparent #model
- Make the #infrastructure right
- Monitor the right set of #metrics
- System should get a boost with a simple model itself
- Launch the product without #MachineLearning . Collect #Data
Phase 1
- Building the first basic/transparent #model
- Make the #infrastructure right
- Monitor the right set of #metrics
- System should get a boost with a simple model itself
Phase 2
- More #FeatureEngineering on top of simple model (Low hanging fruits). #Regularization.
- Multiple launches with new features
- Metrics should still be rising with each launch
- More #FeatureEngineering on top of simple model (Low hanging fruits). #Regularization.
- Multiple launches with new features
- Metrics should still be rising with each launch
