Well if you’re pondering beginning an artificial intelligence (AI) software startup you need to move quick, and your future infant unicorn will require hardware and software. For AI hardware Nvidia is the marker leader and for software you need to finalize which m/c learning software you need to best utilize.
These days you are not going to re-invent the wheel yet rather use an existing platform to build up your solution on. In all likelihood you’ll need to utilize an “open source” platform that you can download now below and kick it into high gear. What open source means is that the product is allowed to utilize and you’ll be joining a whole group of individuals who are creating on the platform simply like you. They’ll likely be a ton more astute than you, yet don’t stress.
Here are the rundown of open source AI Tools:
|Name of Tool||Company||Remarks|
Where to download:
|Best and number one machine learning framework .Google chosen to open source their counterfeit neural system and it’s called Tensorflow.|
H2O Where to download:http://www.h2o.ai/download/
|The best machine learning platform on Spark and Hadoop|
Where to download:
|Not-for-benefit||Bolstered by gifts from Nvidia and Amazon. Caffe can prepare around ~50 million pictures for each day with a solitary NVIDIA GPU.|
|Torch||Not-for-benefit||Utilized inside Facebook, Google, Twitter. Figuring system for machine learning and PC vision.|
|Theano||Not-for-benefit||A Python library that enables to assess numerical expressions including multi-dimensional clusters.|
|scikit||Not-for-benefit||Machine learning in Python. On the off chance that you need to see heaps of cool cases watch here http://scikit-learn.org/stable/auto_examples/index.html.|
|Neon||Nervana Systems||The “world’s speediest deep learning framework”. Obtained by Intel in last year ago ,they’ll up-sellyou to utilize their cloud once you build your application on their platform.|
|Deeplearning4j||Skymind Startup||Deep learning library for Java that needs to be the “Red Hat of Deep learning”.|
|CNTK||Microsoft||Microsoft’s “subjective toolbox” that trains and assesses Deep learning algorithms with langs like C++ ,Python. Quick and versatile. Upgraded for their cloud platform Azure.|
|Mahout||Apache Software||Quickly make versatile machine learning applications. Core algorithms for clustering, collaborative filtering, and characterization.|
|NuPIC||Numenta||System roused by neuroscience called “Hierarchical Temporal Memory” or HTM. Naturally motivated machine insight.|
|OpenNN||Artelnics||Started by a Spanish startup called Artelnics. An open source class library written in C++ programming dialect which actualizes neural systems.|
|PredictionIO||Salesforce||Open source machine learning stage for building predictive applications . Described as “MySQL for machine learning”.|
|Seldon||Seldon||London startup established by information researchers. Full-stack open-source machine learning answer for proposal motors.|
There might be few more tools I will list down ASAP.
“Software is eating the world, but AI is going to eat software.” – Jensen Huang – Nvidia CEO