Azure Machine Learningの評判・口コミ 全21件

What do you dislike?

It is sometimes hard to figure out which API might fit best in your current project, e.g. the handwriting API ist sometimes best to use printed text

What do you like best?

If you wnat to start with Machine learning, the APIs provided make it really easy for you. Even advanced features as well as custom models are well documented and easily created

What do you dislike?

I dislike the price point at which Azure markets itself. It does not make sense to charge that much given their lower costs compared to competitors.

What do you like best?

I appreciate the ability for azure to add in dimensionality to processing power. It is extremely power as a PAAS SAAS IAAS platform and allows users to determine on their own the best approach to the solution of their specific individual needs. I also like the quality with which Azure provides customer support and help assistance to clients. It is a robust platform with a robust user base.

What do you dislike?

It can seem as there is a limited amount of features

What do you like best?

The best feature is the fact it is easy to use and navigate to use the different features

Recommendations to others considering the product

Use this software it is easy to use and get used to, you will find most of your duties can be sorted with this software

What do you dislike?

The formatting of the files can be difficult to categorize.

What do you like best?

This data storage site makes it easy to store large files needed in computer data analyses.

What do you dislike?

The only pitfall to using this form of technology is that it is not the easiest to understand at first. I had to maneuver through the tabular features before I could move forward.

What do you like best?

Azure learning machine is an advanced form of technology that has allowed to crest datasets and view them in an highly organized manner.

What do you dislike?

still evaluating the product , will need to come up soon.

What do you like best?

I like Spark functionality - as real time data stream Azure helps in deep machine learning.

Serverless, drag and drop development - Studio is awesome.

What do you dislike?

So far, I don't find anything I dislike about Azure Machine Learning, but I must admit that my level of apprentice in this field is at a beginner's level.

What do you like best?

First of all, I must say that its interface is so pleasant to the eye and searching and dragging in the different components into the dashboard is fun. Configuring the modules is very simple, with a lot of features being preconfigured. Visualizing the results is also very neat and being able to publish the service as web API with minimal effort is so cool.

Recommendations to others considering the product

An intermediate to advanced knowledge of regression models, data sorting and cleansing, is a must, as with other machine learning software. Also comparing the different learning models, one another is definitely necessary to optimize the results.

What do you dislike?

Not much to dislike - I'm very impressed with the studio capabilities.

What do you like best?

The automated processing of selected algorithms is impressive.

What do you dislike?

Data Shaping/ Data cleaning interface is clunky at the best

Trying to upgrade the package from trail gives you no idea about the subscriptions available.

Its simply not seemless

What do you like best?

1. its easy it to create a web service and then deploy it to production

2. Web Based clean user interface

Recommendations to others considering the product

Be ready to have custom code and get your hands dirty, there is no way out of having your own deep learning cluster in house and on cloud with GPU (if not multi gpu then at least single gpu). TensorFlow has the highest visibility and resource availability in market currently