What is Data Science - Data Science Topics (2)

Data Science Topics

Deep Learning and Machine Learning

Artificial Intelligence and Data Science

Big data

• massive, quickly built, so varied datasets
• organizations now have the power to analyze these vast data sets
• velocity, volume, variety, veracity, value (<- the 5Vs from the previous lecture)

Data Mining

• process of automatically searching and analyzing data, discovering previously unrevelaed patterns
• involves preprocessing -> transforming -> data visualizations

Machine Learning

• subset of AI that uses computer algorithms to analyze data
• make intelligent decisions based on what it has learned, without being explicitly programmed
• trained with large sets of data, learn from examples, do not follow rules-based algorithms
• enables machines to solve problems on their own and make accurate predictions with provided data

Deep Learning

• specialized subset of machine learning that uses layered neural networks to simulate human decision-making
• deep learning algorithms can categorize information and identify patterns
• enables AI systems to continuously learn on the job and improve the quality of accuracy of results

Nerual Networks

• take inspiration from biological nerual networks, although they work quite a bit differently
• collection of small computing units called neruons that take incoming data and learn to make decisions over time
• often layer-deep and are the reason deep learning algorithms become more efficient as the data sets increase in volume, as opposed to other machine learning algorithms that may plateau as data increases

Between Aritificial Intelligence and Data Science

• Data science is the processs and method for extracting knowledge and insights from large volumes of disparate data
• involves mathematics statistical analysis data visualization machine learning and more...
• could use machine learning algorithms, deep learning models to extract meaning and else
• broad term encompasses the entire data processing methodology (encompass : 포함하다, 아우르다)

AI includes everything that allows computers to learn how to solve problems and make intelligent decisions.

Both AI and Data science can involve the use of Big Data.

Generative AI and Data Science

What is Generative AI?

• subset of ai, produces new data
• allows machine to create content such as images, music, language and code

How does Generative AI work?

• Generative Advisarial Networks GANs
• Variational Auto-Encoders VAEs
• Replicates underlying features of original data

Applications of Generative AI

• Natural Language Processing like GPT-3
• Healthcare
• Art and Design
• Gaming. realistic environments
• assits fashion and retail with shopping recommendations

Syntehtic Data (synthetic : 합성한, 인조의, 종합적인)

• building data models takes a lot of data
• data sets may not have enough data to build a model
• generative AI makes data augmentation possible
• creates data with similar properties (distribution, clustering)
• use this synthetic data for model training and testing. examining data patterns inside

Coding Automation

• confined by a time when examining data (confine : 국한시키다, 가두다, 얽매이다)
• generate software code to construct models
• focus on higher-level tasks

Uncover insights

• generate insights and reports
• automate updates
• enhance decision-making
• IBM Congos Analytics
• Music, language, code, ...

Recap

• Gerative AI, a subset of artificial intelligence, focuses on producing new data rather than analyzing existing data
• Gernative AI augments data science efforts, enabling deeper insights, addressing data limitations, and improving the overall quality of data-driven outcomes

Nerual Networks and Deep Learning

• neural networks try to mimic human brains
• keep feeding the inputs in to check what kind of transformation happen between inputs and outputs

**

또 교수님의 한 마디...

그냥 수업 시간 중 교수님의 수다같은 느낌으로 듣고 있다...

**

Applications of Machine Learning

• predictive analytics
• recommendation systems...
• fraud detection

**

얘도 마찬가지...

그냥 교수님 훈화말씀 같은 느낌...

**

Regression

The department of obvious conclusions

• 그냥 글쓴이의 주절주절
• 큰 집이 더 비싸다!! 뿐이 아니라 화장실 개수에 따른 가격상승률이 방 개수보다 크다
• freeway나 highway 주변에 있는 집들이 다른 집들에 비해 더 적게 팔린다
• 쇼핑 센터랑 엄~청 가까운 집들보다는 좀 적당히 거리가 있는 집들이 더 많이 팔린다

Why Regress (hedonic)?

• 아래 질문들에 대해 상관관계를 한 번 생각해보렴
• How much more can a house sell for an additional bedroom?
• What is the impact of lot size on housing price?
• Do homes with brick exteriors sell for less than homes with stone exteriors?
• How much does a finished basement contribute to the price of a housing unit?
• Do houses located near high-voltage power lines sell for more or less than the rest?

Practice Quiz : Regression

Q. What are some examples of questions that can be addressed using regression (hedonic) models in the context of housing prices?
A. How much does the size of a lot influence housing prices?

 

Q. What is the primary purpose of regression hedonic models in the context of housing analysis?
A. To analyze the relationships between various factors and housing prices

 

Q. You are ready to buy a house. However, you wonder, "Do houses located near high-voltage power lines sell for more or less than the rest?" This question can be addressed using regression analysis.
A. True

Lesson Summary : Deep Learning and Machine Learning

AI Terminology

• AI terms you should know : machine learning, deep learning, neural networks, generative AI

Artificial Intelligence

• Branch of computer science
• devlopment of systems that can mimic many of the tasks with human intelligence

Machine learning

• using computer algorithms to analyze and make predictions
• without needing to explicitly program the analysis methods into the System

Deep Learning

• subset of machine learning that uses layered nerual networks to simluate human decision-making

Nerual Network

• collection of small computing units, called neruons, that take incoming data and learn to make decisions over time
• more efficient as the amount of data increases in volume, unlike other machine learning algorithms, which tend to plateau

Generative AI

• subset of AI
• focus on new data productions
• mimic content created by humans
• can generate new data to use when training and testing a model

Machine Learning algorithms

• applications : make predictions, make recommendations, ...

Regression

• identifies correlation between inputs and ooutputs
• ex. predict the cost of a house based on square footage and number of bathrooms

Generative AI procduces new data from raw data with similar traits. (trait : 특성)
Deep learning uses nerual networks to teach iteself patterns from inputs and outputs.
Data scientists use AI along with "big data" to make predicitons.

Practice Quiz : Deep Learning and Machine Learning

Q. Imagine you're working on an AI project that involves creating new content such as images, music, and language, Which artificial intelligence technology would you.be primarily focused on?
A. Generative AI

 

Q. What sets deep learning apart from traditional nerual networks?
A. Multiple layers of nerual networks

 

Q. In the realm of machine learning, what significant application involves the task of predicting items of interest for users based on their past interactions or behaviors?
A. Recommender systems for personalized content suggetsions.