scikit-learn Reviews & Product Details

The best aspect about this framework is the availability of well integrated algorithms within the Python development environment. It is quite easy to install within most Python IDEs and relatively easy to use as well. A lot of tutorials are accessible online which supplements understanding this library allowing to become proficient in machine learning. It was clearly built with a software engineering mindset and nevertheless, it is very flexible for research ventures. Being built on top of multiple math-based and data libraries, scikit-learn allows seamless integration between them all. Being able to use numpy arrays and pandas dataframes within the scikit-learn environment removes the need for additional data transformation. That being said, one should definitely get familiar with this easy to use library if they plan on becoming a data-driven professional. You could build a simple machine learning model with just 10 lines of code! With tons of features like model validation, data splitting for training/testing and various others, scikit-learn's open source approach facilitates a manageable learning curve. Review collected by and hosted on G2.com.

One issue that has persisted and troubled me since quite some time is the lack of categorical variables transformation capabilities (it is much easier in libraries like tensorflow). It is comparatively slower than tensorflow when it comes to big datasets and this is something that should be adopted soon especially in the era of big data technologies. However, with the frequency of updates, I believe most issues get resolved really quickly making it a robust package for machine learning development. Review collected by and hosted on G2.com.

It is the platform machine learning, easy to learn, easy to test

provides all the capability that any machine learning platform should have, lot of algorithms like encoders - binary encoder, one hot encoder

provides implementation for all supervised and un-supervised learning

provides all the ability to validate the model

we can integrate easilty with mat plotlib, pandas, numpy and for serialisers

lot of specific example tutorials in internet available for the beginners

It is open source and totally free

lot of the other open source and many propriatry products for ml are developed on top of the sci kit library

as it provides python interface easy to learn and integrate with many other platforms Review collected by and hosted on G2.com.

there are two problems which i can mention are

1. not possible to scale horizontaly

2. have issues when we have categorical attributes in variables - encoding them to binary or one hot encoded will not solve the issue

Many of the recent technologies like h20, tensor flow gives the ability to inout categorical attributes as inputs to algorithm Review collected by and hosted on G2.com.

It is very strong tool being used in data science especially in machine learning. It is open source and free package that has great role in machine learning. It has great ability that we can integrate it with other packages such as mat plotlib, nympy and pandas. It has a great role in data science and machine learning algorithms. Review collected by and hosted on G2.com.

It has great features. However it has some drawbacks dealing with categorical attributes. Otherwise it is very strong package. I do not see any other drawbacks of using this package. Review collected by and hosted on G2.com.

Amazingly useful tool set for machine learning and data science work. Personally use it in python and it's really helpful. Some popular package such as pandas, numpy and matplotlib add it even more values. I always use it besides neural networks and yield solution as a combination, and the solution gives the best result often comes from it, by working on different points. Review collected by and hosted on G2.com.

No, nothing comes to my mind for dislike part and I have used it for couple years in machine learning competitions and projects. They also update scikit-learn quite often to fix any known issue and make improvement. Review collected by and hosted on G2.com.

I like the fact that it includes a ton of functionalities and incorporates almost all of the Machine Learning algorithms meant for supervised and unsupervised learning.

It can be used to develop various regression, classification and clustering algorithms.

It utilizes a range of machine learning, preprocessing, cross-validation and visualization algorithms.

It provides three Regression Metrics namely Mean Absolute Error, Mean Squared Error, R² Score.

It also provides three Classification Metrics namely Accuracy Score, Classification Report, Confusion Matrix.

Additionally, it provides three Clustering Metrics namely Adjusted rand Index, Homogeneity, V-measure. Review collected by and hosted on G2.com.

Apart from the inability to scale well, there’s also the fact that scikit-learn does absolutely nothing to assist with deploying the model to production. Review collected by and hosted on G2.com.

- It contains many machine learning algorithms such as: random forest, decision tree, support vector machines, linear discriminant analysis, quadratic discriminant analysis, logistic regression, multi layer perceptron(neural networks), naive bayes, other boosting algorithms, knn, k-means (and other clustering algorithms)

- It contains preprocessing tools (normalization, standardization)

- It contains hyperparameter tuning tools (RandomSearchCV, GridSearchCV)

- It contains many kinds of metrics to tune the model for (accuracy, recall, precision, f1_score, etc)

and summing up all these it is possible to develop and create an end to end machine learning application

Not to mention all of these above along with scikit-learn as a whole are compatible with other Python libraries such as pandas, numpy, mlxtend, matplotlib. Review collected by and hosted on G2.com.

- It should include more recent state of the art algorithms such as XGBoost, Catboost, LightGBM.

- It should facilitate GPU, otherwise hyperparameter tuning takes too much time Review collected by and hosted on G2.com.

Various machine learning models and easy to adjust parameters. Also easy to apply data transformation prior to fit the model Review collected by and hosted on G2.com.

Could add more examples in the documentation Review collected by and hosted on G2.com.

1. I love the fact that I can try out a variety of machine learning algorithms without having to build them from scratch. I just call them using functions already available.

2. Scikit-learn provides users with a function to split a given dataset into train and validation data by just passing a split ratio only.

3. Scikit-learn easily integrates with other deep learning frameworks. Review collected by and hosted on G2.com.

I have not had any reason to hate scikit-learn at the moment, as it has helped me achieve a lot in machine learning. Review collected by and hosted on G2.com.

Scikit learn is simply wonderful. It abstracts away all the complexities of several machine learning frameworks. Scikit learn provides beautiful one line function calls to really complex functions and the documentation is beautiful. A complete noob can go through their documentation and understand since it is human readable. In addition to top machine learning models ranging from random forest, decision trees and linear regression, they also provide libraries for data preprocessing. You can do data preprocessing, one hot encoding and lots of other things with Scikit Learn. Review collected by and hosted on G2.com.

Scikit learns models take a long time to train, and they require that your data be in a specific format. This can be really stressful especially when the error messages don't provide much insight into the problem Review collected by and hosted on G2.com.

I like this library because it is super easy to import the library and use the Machine Learning models.

To install scikit-learn it is very easy.

They have lots of machine learning models such as random forest, xgboost and many more. You don't need to code from scratch. They provide a lot of parameters to tweak the models also which is helpful. Review collected by and hosted on G2.com.

It is kind of plug and plays but the customization is a little bit hard for the machine learning models. Also, as compared to tensorflow it is slow. Review collected by and hosted on G2.com.