Skip to main content

AI for Everyone

You might have heard of a lot about AI, Machine learning, Data science, Deep learning, etc,... But, what exactly these terms mean and how is the connection between those. Here is my understanding:

There are two ideas of AI:
- ANI (Artificial Narrow Intelligence): E.g., smart speaker, self-driving car and web search.
- AGN (Artificial General Intelligence): do anything a human can do.

ANI is realistic and incredibly valuable. Though AGN is still too far away, and there is no need to unduly worry about it.

When talking about data in term of AI that means talking about dataset. There are several methods to get a dataset:
- Manual labeling
- Observing user behaviors
- Observing behaviors of other things such as machine
- Downloading dataset from a website or acquiring it from a partner.

Machine learning (ML) is a tool in AI. Supervised learning is a type of ML that learns A to B, or input to output mappings. Deep learning/Neural networks is a type of supervised learning which can maximize the performance of AI when dataset is big today.

Is AI a subset of data science?

No! Data science uses many tools from AI machine learning, but has some other separate tools as well that solves a very set of important problems in driving business insights.

- ML project output: a running AI system
- Data science project output: a set of insights


Reference
Labels:

Comments

Post a Comment

Popular posts from this blog

Google I/O 2017 Notes

WOW! How meaningful this below video explains about the name of  "I/O". Sundar Pichai talked a lot of Machine Learning Machine Learning is a very hot trend these days. Google uses it for their products. Google Assistant: Easily booking an online meal by talking with Google Assistant like a staff of partners, for example. Google Home: Hands-free calling. Google Photos: sharing suggestion, shared library, photo books and google lens. Youtube: 360 degree video, live stream. Kotlin became an official programming language for Android https://kotlinlang.org I'm on the way to Kotlin! ^^ Reference: [1]. https://www.youtube.com/watch?v=Y2VF8tmLFHw
Post a Comment
Read more

Junit - Test fails on French or German string assertion

In my previous post about building a regex to check a text without special characters but allow German and French . I met a problem that the unit test works fine on my machine using Eclipse, but it was fail when running on Jenkins' build job. Here is my test: @Test public void shouldAllowFrenchAndGermanCharacters(){ String source = "ÄäÖöÜüß áÁàÀâÂéÉèÈêÊîÎçÇ"; assertFalse(SpecialCharactersUtils.isExistSpecialCharater(source)); } Production code: public static boolean isExistNotAllowedCharacters(String source){ Pattern regex = Pattern.compile("^[a-zA-Z_0-9_ÄäÖöÜüß áÁàÀâÂéÉèÈêÊîÎçÇ]*$"); Matcher matcher = regex.matcher(source); return !matcher.matches(); } The result likes the following: Failed tests: SpecialCharactersUtilsTest.shouldAllowFrenchAndGermanCharacters:32 null A guy from stackoverflow.com says: "This is probably due to the default encoding used for your Java source files. The ö in the string literal in the J...
5 comments
Read more

JSF, Primefaces - Invoking Application Code Even When Validation Failed

A use case I have a form which has requirements as follow: - There are some mandatory fields. - Validation is triggered when changing value on each field. - A button "Next" is enable only when all fields are entered. It turns to disabled if any field is empty. My first approach I defined a variable "isDisableNext" at a backend bean "Controller" for dynamically disabling/enabling the "Next" button by performing event "onValueChange", but, it had a problem: <h:form id="personForm"> <p:outputLabel value="First Name" for="firstName"/> <p:inputText id="firstName" value="#{person.firstName}" required="true"> <p:ajax event="change" listener="#{controller.onValueChange}" update="nextButton"/> </p:inputText> <p:outputLabel value="Last Name" for="lastName"/> <p:i...
Post a Comment
Read more

There is no expert, there is only us

Chairs in my office Once I was chatting with my teammates... Me : "I wonder why it is always necessary to refer to this source, that source, has anyone done it, is there any research? While no problem is the same, no situation is the same. Why not use our own brains to create a whole new one? Why do we have to be the insiders to consult the outsiders? Isn't that very paradoxical?" My teammate : "I agree with your point that no two situations will be exactly the same and that there is no one size fits all. But there's one like this, I don't think it's all the problems we're dealing with that only we can meet. So of course, if I can find a source to refer to first, it's still better. Just like when implementing a software feature, not every problem is the same. But I still have to go check to see if I'm the only one I've met? Or if someone has met me, I have to see if their method suits me before I decide to try it, but I don't blindly ...
Post a Comment
Read more