Pullquotes

Rozina Myoya · December 11, 2019

HTML Code

In graphic design, a pull quote (also known as a lift-out pull quote) is a key phrase, quotation, or excerpt that has been pulled from an article and used as a page layout graphic element, serving to entice readers into the article or to highlight a key topic. It is typically placed in a larger or distinctive typeface and on the same page. Pull quotes are often used in magazine and newspaper articles, annual reports, and brochures, as well as on the web. They can add visual interest to text-heavy pages with few images or illustrations.

Placement of a pull quote on a page may be defined in a publication’s or website’s style guide. Such a typographic device may or may not be aligned with a column on the page. Some designers, for example, choose not to align the quote. In that case, the quotation cuts into two or more columns, as in the example shown. Because the pull quote invites the reader to read about the highlighted material, the pull quote should appear before the text it cites and, generally, fairly close to it.

Pull quotes need not be a verbatim copy of the text being quoted; depending on a publication’s house style, pull quotes may be abbreviated for space or paraphrased for clarity, with or without indication.

A disadvantage of pull quotes as a design element is that they can disrupt the reading process of readers invested in reading the text sequentially by drawing attention to ghost fragments out of context. At the other extreme, when pull quotes are used to break up what would otherwise be a formless wall of text, pull quote can serve as visual landmarks to help the reader maintain a sense of sequence and place.

Fork this repository, then rename the repository to yourgithubusername.github.io.

Alternatively, you can use Use this template button if you want to create a repository with a clean commit history which will use Reverie as a template.

Your Jekyll blog will often be viewable immediately at https://yourgithubusername.github.io (if it’s not, you can often force it to build by completing step 2)

Enter your site name, description, avatar and many other options by editing the _config.yml file. You can easily turn on Google Analytics tracking, Disqus commenting and social icons here.

Making a change to _config.yml (or any file in your repository) will force GitHub Pages to rebuild your site with jekyll. Your rebuilt site will be viewable a few seconds later at https://yourgithubusername.github.io - if not, give it ten minutes as GitHub suggests and it’ll appear soon.

Create a new file called /_posts/2019-2-13-Hello-World.md to publish your first blog post. That’s all you need to do to publish your first blog post! This Markdown Cheatsheet might come in handy while writing the posts.

You can add additional posts in the browser on GitHub.com too! Just hit the Create new file button in /_posts/ to create new content. Just make sure to include the front-matter block at the top of each new blog post and make sure the post’s filename is in this format: year-month-day-title.md

You can categorize your content based on categories in Reverie. For this, you just need to add categories in front matter like below:

For adding single category:

categories: JavaScript

For adding multiple categories:

categories: [PHP, Laravel]

The contegorized content can be shown over this URL: https://yourgithubusername.github.io/categories/

RSS

The generated RSS feed of your blog can be found at https://yourgithubusername.github.io/feed. You can see the example RSS feed over here.

Sitemap

The generated sitemap of your blog can be found at https://yourgithubusername.github.io/sitemap. You can see the example sitemap feed over here.

License

MIT

The approach of this post is to used the kNN mode seeking ensemble method on arbitrary data and measure the performance, while commenting on the

Financial intitutions invest enourmous amounts of money running campaigns on new product offerings to their existing customers and to acquire new business through traditional media and (most recently) digital media strategies. The use of these marketing strategies is to determine the customer segments that would be most likely to subscribe to their product offerring.

To determine these different customer segments, statistical analysis is usally carried out on historical data collected by the financial instituation (such as a bank), and statisticians and/or data scientist attempt to predict the likelihood of a customer (with a certain combination of attributes) subscrbing to a product. These results then inform the most appropriate marketing strategy that should be followed to maximise customer subscription.

Thus bringing us to the example of a marketing campaign ran by a Portuguese banking institution that had the goal of making predictions on whether an exisiting customer would subscribe to the “bank term deposit” product they were offering. The predictions made involved carrying out a Logistic Regression analysis taking into account the significant factors that would influence the probability of a customer saying “yes” to the product subscription. The marketing campaigns were based on phone calls (either cellular or telephone) and often multiple contact sessions with the client was necessary to assess the client’s receptivity to subscribing to the “bank term deposit” product.

Development of a proposed marketing strategy involved:

Feature selection process to determine the significant variables to be used in the regression process
Logistic regression and model performance evaluation to determine whether reliable predictions can be made using the estimated model
Insight derivation and conclusions on the most appropriate marketing strategy to be followed

Feature Selection

The dataset can be found here: Dataset.

The total data set consists of 50 422 observations with 16 attributes (including the response variable). The data set was split into two sets; one set was used for the regression process, while the other was used to conduct the predictions. The two sets will be referred to as the “Regression set” and the “Prediction set” respectively. The regression set consisted of 45 211 observations and the prediction set consisted of 5 211 observations.

The list of varibles considered were:

#	Variable label	Variable Description
1	age	Numeric: Age of the client
2	job	Categorical
3	marital	Client’s marital status
4	education	Education level achieved
5	default	Has credit in default?
6	housing	Has a housing loan?
7	loan	Has a personal loan?
8	contact	Contact communication type
9	month	Month of the year (Jan to Dec)
10	campaign	umber of contacts performed during the campaign and for the specific client
11	pdays	Number of days that passed by after the client was last contacted from a previous campaign
12	previous	Number of times the client had been contacted before the campaign
13	poutcomes	Outcomes of previous marketing campaigns
14	y	Did the client subcribe a term deposit? (yes or no)

The data consisted of both numeric (continuous) variables and categorical variables, therefore, to determine whether there was any correlation between the attributes the following three metrices were used:

Contingency Coefficient ($\mathcal{C}$) - Categorical Data “coerrelation”
One way ANOVA test - Categorical and Numerica data “correlation”
Pearson’s correlation coefficient - Numeric data correlation

Given the features were not correlated, I went forward with the Logistic regression process to estimate a model.

Making Predictions

Step 1) Variable Selection and Model Development