Top Rated MLlib Alternatives
13 out of 14 Total Reviews for MLlib
Overall Review Sentiment for MLlib
Log in to view review sentiment.
It is useful in implementing machine learning algorithms like classification, regression and clustering. It works well while using statistical modelling techniques Review collected by and hosted on G2.com.
It has an expensive memory with the necessity of manual optimization which might degrade user experience. It gives latency but can be used amongst R and python communities Review collected by and hosted on G2.com.
implementation of ML algorithms like regression, classification and modelling techniques can be done using the tool Review collected by and hosted on G2.com.
MLlib is not production ready, moreover Spark does not come out as a useful engine owing to its latency Review collected by and hosted on G2.com.
The scalability power of the framework which handles large data efficiently and performs machine learning algorithms at faster rate. Review collected by and hosted on G2.com.
The syntax and code changes for python R depends on the tools we are using.It is not standard which is tough for new users to adapt.The packages are very different compared tools to tool. Review collected by and hosted on G2.com.
MLLib was used as part of course in my college for Big Data. So we got to study why actually mllib came about and what all inadequacies were there in the Map-Reduce Framework of Hadoop and how apache Spark has solved them. The best part is the ease of use of Mllib and also the excellent documentation support from both the official website as well as the sources outside like youtube videos. The big community makes it easy to learn and use mllib. I used mllib for decision trees and I being a student was successfully able to implement the same with ease. Plus the python implementation is very easy to implement. Review collected by and hosted on G2.com.
We were given a preinstalled system for our labs and a cluster, but when I tried to do the same for my machine, I found it rather tricky to install. Also, support for deep learning is not there, which is a very fast growing field of machine learning. Review collected by and hosted on G2.com.
MLib so far is the best community supported widely used machine learning library for apache spark Review collected by and hosted on G2.com.
MLib is inconsistent with deep learning models, this causes issues while moving models to production Review collected by and hosted on G2.com.
Speed and ease of use. Strong community support and lots of resources. Review collected by and hosted on G2.com.
Prototyping can be time consuming. Also, limited utility in case of extremely large datasets. Review collected by and hosted on G2.com.
I love how it includes most of the popular ML libraries for easy use with Apache Spark and parallelized computing. The power is only limited by the number of cores you've got. Review collected by and hosted on G2.com.
I feel like some other ML frameworks have better models, or added features/functionality used in developing models. MLlib is also an open source part of Spark, so development of the framework depends largely on what Open Source folks contribute to it. Review collected by and hosted on G2.com.
It is distributed and allow distributed execution of model training ans well as model scoring. It helps to leverage benefit of Spark without using Scala. It delivers Spark ML with Python!
High performance since it is a RDD-based data modeling package.
Fairly nice documentation. Review collected by and hosted on G2.com.
It is rigid with some of the algorithms, specially with advanced one like neural network. For instance, you are unable to change activation functions of a neural network. You can either use Sigmoid for all the layers, or tanh which is not really making sense!
Evaluation metrics are not as rich as packages like Scikit-Learn.
Not all its functionalities implemented in Python. Many are Scala-based yet. Review collected by and hosted on G2.com.
