VIP Class Notes (Raph) [W]

Vocabulary

Intensive (adj): 1- involving a lot of effort or activity in a short period of time.
E.g.: All new workers need to go through two weeks of intensive training
E.g.: She’s taking an intensive course in English this summer.

Incentive (n): 1- something that encourages a person to do something.
E.g.: There is little incentive for people to leave their cars at home when public transport is expensive.
E.g.: Bonus payments provide an incentive to work harder.

Overkill (n): 1- much more of something than is needed, resulting in less effectiveness.
E.g.: Should I add an explanation, or would that be overkill?
E.g.: I don’t know if that’s necessary or if it’s a little bit of overkill.
E.g.: Going for a milder perfume is always best, to avoid overkill.

Grammar

Firstly, the students need to watch the video uploading in the website. – Firstly, the students need to watch the video uploaded to/on the website.

She drop her work and study whole day. – She quit her job and studies/is studying full-time.

I need to find many datas. – I need to find a lot of data.

Writing exercise

Original:

1. In some retail company, there are two product lines, which are Full-Price product and Outlets product. Every kind of product line focus on different types of customers, who have different characteristics such as income per year, buying behavior etc. There is a customer database and sales records. The product type is used as observation value, which records the type that customer bought most for the last year. Support vector machine can be used as classification model. A classifier can be used to separate people who buy different type of products. The company can use the result to tag every upcoming customer and send some target commercial advertisement to them.

2. An investment bank might investigate many companies’ financial situation and create a database. The dataset might include year when the company was founded, revenue, the number of employees and profit etc. The bank can select a subset, which includes companies being founded for three years. Observation values could be whether the company has a positive profit. The Support Vector machine can also be used as classification model. A training subset is used for finding a proper classifier and a testing subset is used to validate it.

3. A school might try to give high school students suggestion that which rank of university can be applied based on their GPA, performance, activities and awards earned etc. Graduate information can be used for analysis. The respond can be defined as a certain level of universities or colleges. For example, 1 might represented the universities or colleges ranking in top 50. K-nearest-neighbors model might be used for this classification. The school can use the training subset to find a proper model and use the testing subset to find how the model is fit for.

Edited:

1. In some retail companies there are two product lines: Full-Price and Outlet. Each product line focuses on different types of customers, who have different characteristics, such as income per year, buying behavior etc. Those companies can apply “product type” as an observation value in their customer databases and sales records, and record what each customer bought most for the last year. Support vector machine can be used as a classification model. A classifier can be used to separate people who buy different types of products. The company can use the results to classify upcoming customers and send them target advertisement.

2. Investments could benefit from using classification models, such as the Support Vector Machine. In such situation, an investment bank might be interested in investigating the financial situation of many different companies and creating a database. The dataset might include year when the company was founded, revenue, the number of employees, profit margin, etc. The bank can select a subset, which includes companies that have been operating, for example, for three years. Observation values could be whether the company has a positive net income. The Support Vector machine can also be used as classification model. A training subset is used for finding a proper classifier and a testing subset is used to validate it.

3. A school might want to advise their high school students on what universities they might have the best chances of a successful application, based on their GPA, other performance measures, extracurriculars, awards earned etc. Information on past students can be used for analysis. The respond can be defined as a certain level of universities or colleges. For example, 1 might represented universities or colleges ranking in top 50. K-nearest-neighbors model might be used for this classification. The school can use the training subset to find a proper model and use the testing subset to find how appropriate the model is.

Pronunciation

Model: /ˈmÉ‘Ë.dÉ™l/

Accuracy: /ˈæk.jɚ.ə.si/

Criteria: /kraɪˈtɪr.i.ə/

Zone:  /zoʊn/