FIT1043-Week1.1-DataScience.pdf

FIT1043 Introduction to Data Science Week 1, Data Science Dr. Sicily Ting Fung Fung School of Information Technology Monash University Malaysia  Overview of Data Science ePub Section 1.1 Learning Outcomes Week 1 By the end of this week you should be able to: Explain what is data science an...

FIT1043 Introduction to Data Science Week 1, Data Science Dr. Sicily Ting Fung Fung School of Information Technology Monash University Malaysia  Overview of Data Science ePub Section 1.1 Learning Outcomes Week 1 By the end of this week you should be able to: Explain what is data science and Drew Conway’s Venn diagram Comprehend the usefulness of machine learning Explain different components of a data science process Differentiate data science from other related disciplines Learn how to install and start coding in Python with Jupyter Notebook To be achieved in your tutorial / laboratory session Data Scientist Different professions What is Data Science? “name contains the word ‘science’, so it can’t be one” This is a very old joke as most of us know sciences as Biology, Physics, Chemistry and so on. They don’t have the word science in them ☺ “data science is what a data scientist does” A circular definition, which is practically useless. Please do not answer like this. What is Data Science? “data science is the technology of handling and extracting value from data” This is a less circular definition and slightly more useful. It isn’t a bad attempt at it. “machine learning on big data” Useful, but it is too narrow. What is Data Science? Data Science and Big Data (links to Wikipedia) Data Science Data Science is the extraction of knowledge from data, which is a continuation of the field data mining and predictive analytics. Data science is an inter-disciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from structured and unstructured data. Data science is related to data mining and big data. Big Data Big data is a broad term for data Big data is a field that treats ways to sets so large or complex that analyze, systematically extract traditional data processing information from, or otherwise deal applications are inadequate. with data sets that are too large or complex to be dealt with by traditional data-processing application software. What is Data Science? Quote from Hal Varian The ability to take data and; to be able to understand it, to process it, to extract value from it, to visualize it, to communicate it That’s going to be a hugely important skill in the next decades. What is Data Science? In Summary Examples narrow: machine learning on big data broad: extraction of knowledge/value from data through the complete data lifecycle process Should include broad concern with the focus on the different stages learning/knowledge discovery  http://drewconway.com/zia/2013/3/26/the-data-science-venn-diagram Data Science Venn Diagram Drew Conway (http://drewconway.com/) Data Science Venn Diagram Able to obtain the data (using your hacking skills) plus mathematics & statistics gets you machine learning. Many CS graduates starts with the title Machine Learning Engineer Substantive expertise plus mathematics & statistics is traditional research. Doctoral level researchers spend a lot of their time acquiring expertise in specific areas, but very little time learning about technology. Capable of extracting and structuring data and has good knowledge of the field equates a danger zone. Appears to be a legitimate analysis without any understanding of how they got there or what they have created Data Science Venn Diagram Conclusion Combination of different skill sets Diverse skills are needed Data Science Applied Examples of Data Science being applied Microsoft predictive analytic for traffic forecast iOS predictive text Google translation engine Amazon recommender system Health research studies on saturated fat Machine Learning Machine Learning Definition Well understood definition and widely agreed upon: Machine Learning is concerned with the development of algorithms and techniques that allow computers to learn. Concerned with building computer programs that can learn, often with computational output. With the underlying theory in statistics Further reading at A Gentle Guide to Machine Learning Why Use Machine Learning? Machine learning is useful when: Human expertise is not available, e.g. Martian exploration Humans cannot explain their expertise (as a set of rules), or their explanation is incomplete and needs tuning, e.g. speech recognition Many solutions need to be adapted automatically, e.g. user personalisation image sources: theconversation.com, meduim.com, blog.prioridata.com Why Use Machine Learning? Machine learning is useful when: Situation changes over time, e.g. junk email There are large amounts of data, e.g. discover astronomical objects Humans are expensive to use for the work, e.g. handwritten zipcode recognition image sources: lifewire.com, clealyexplained.com, medium.com  you do not want to be this poor guy! Why Use Machine Learning? sifting through all the data by hand Summary Humans are incapable of it Automation Large Amount of Data Humans are too expensive to be used in such a situation Experts Systems are generally rule based. But many situations change all the time. Recap: Learning Outcomes Week 1 By the end of this week you should be able to: Explain what is data science and Drew Conway’s Venn diagram Comprehend the usefulness of machine learning Explain different components of a data science process Differentiate data science from other related disciplines Learn how to install and start coding in Python with Jupyter Notebook To be achieved in your tutorial / laboratory session http://drewconway.com/zia/2013/3/26/the-data-science-venn-diagram