Financial data

This dataset (.csv) collects 200+ financial indicators for all the stocks of the US stock market.
The financial indicators are found in the 10-K filings that publicly traded companies release
yearly.
The last column of the dataset represent the class of each stock, where:

● if the value of a stock increases during 2018, then class=1;

● if the value of a stock decreases during 2018, then class=0.

In other words, stocks that belong to class 1 are stocks that one should buy at the start of
year 2018, and sell at the end of year 2018.

1. Some financial indicator values are missing (nan cells), so the user can select the
best technique to clean each dataset (dropna, fillna, etc.).
2. There are outliers, meaning extreme values that are probably caused by mistypings.
Also in this case, the user can choose how to clean each dataset (have a look at the
1% - 99% percentile values).
3. The third-to-last column, Sector, lists the sector of each stock. Indeed, in the US
stock market each company is part of a sector that classifies it in a macro-area.
Since all the sectors have been collected (Basic Materials, Communication Services,
Consumer Cyclical, Consumer Defensive, Energy, Financial Services, Healthcare,
Industrial, Real Estate, Technology and Utilities), the user has the option to perform
per-sector analyses and comparisons.
4. The second-to-last column, PRICE VAR [%], lists the percent price variation of each
stock for the year. For example, if we consider the dataset 2018_Financial_Data.csv,
we will have percent price variation for the year 2018 (meaning from the first
trading day on Jan 2018 to the last trading day on Dec 2018).
5. The last column, class, lists a binary classification for each stock, where
○ for each stock, if the PRICE VAR [%] value is positive, class = 1. From a
trading perspective, the 1 identifies those stocks that an hypothetical trader
should BUY at the start of the year and sell at the end of the year for a profit.
○ for each stock, if the PRICE VAR [%] value is negative, class = 0. From a
trading perspective, the 0 identifies those stocks that an hypothetical trader
should NOT BUY, since their value will decrease, meaning a loss of capital.

This dataset has been developed in order to understand whether or not it is possible to
classify the future performance of a stock by looking at the financial information released in
the 10-K filings.
How can you achieve that?

1. Build a classification model for the stocks that would and would not increase their
value in 2018
Big Mart Sales
This dataset contains information collected by BigMart (a supermarket chain in the US)
about sales data for 1559 products across 10 stores in different cities.
The attributes recorded for each product and store have been defined as here below

● Item_Identifier: Unique product ID

● Item_Weight: Weight of product
● Item_Fat_Content: Whether the product is low fat or not
● Item_Visibility: The % of total display area of all products in a store
allocated to the particular product
● Item_Type: The category to which the product belongs
● Item_MRP: Maximum Retail Price (list price) of the product
● Outlet_Identifier: Unique store ID
● Outlet_Establishment_Year: The year in which store was established
● Outlet_Size: The size of the store in terms of ground area covered
● Outlet_Location_Type: The type of city in which the store is located
● Outlet_Type: Whether the outlet is just a grocery store or some sort of
supermarket
● Item_Outlet_Sales: Sales of the product in the particular store. This is
the outcome variable to be predicted.

BigMart has collected such data in order to understand what kind of products sell more in
what kind of stores. Furthermore, it would like to investigate how much the item_visibility
impacts sales. We, as third users, may be interested in segmenting the products according
to their specifics and/or sales at different stores.
How can you achieve that?

1. Build a predictive model and find out the sales of each product at a particular store
(or at generic stores with different characteristics
2. Cluster items according to the available covariates, perhaps considering also the
different sales in different stores (you should spread() the dataset for this last task.
Brazilian Houses
This dataset contains information about 10962 houses to rent in different Brazilian cities. The
data have been gathered through a web-crawler (data have been automatically scraped from
publicly available rent ads in the web), therefore be aware of possible errors or inconsistency
in the data (outliers, duplicates, missing values, etc.).
The following 13 different features have been collected.

● city: City where the property is located

● area: Property area
● rooms: Number of rooms
● bathroom: Number of bathrooms
● parking spaces: Number of parking spaces
● floor: Floor number
● animal: Accept or does not accept animals
● furniture: Furnished or not furnished
● hoa (R$): Monthly Homeowners Association Tax (tassa condominiale),
in Real
● rent amount (R$): Monthly requested rent amount, in Real
● property tax (R$): Yearly property taxes, in Real
● fire insurance (R$): Monthly fire insurance cost, in Real

These data have been collected in order to better understand the house-rent market in some
of the most important cities in Brazil. A new company wants to enter the real-estate market,
and wants to understand what kind of houses grant the larger (rent) revenue before investing
its money: what are the driving forces leading to high rents?
Furthermore, we may want to segment the rent-houses market in different groups: does it
check with the geographical positioning?

1. Build a predictive model and find out the rent amount according to the house
specifics
2. Cluster the houses for rental according to their characteristics.
Garment Workers
This dataset includes important attributes of the garment manufacturing process and the
productivity of the employees which had been collected manually and also been validated by
the industry experts. Data have been collected across different days along the year, and
each row (instance) contains different characteristics of a specific worker team devoted at
performing a specific task.
The following 14 different features have been collected.

● date: Date of the recording

● quarter: Portion of the month (month was divided into 4 quarters)
● department: Department to which the team belong to
● day: Day of the week
● team: Associated team number
● targeted_productivity: Targeted productivity set by the Authority for
the team at that specific day.
● smv: Standard Minute Value, it is the allocated time for a task
● wip: Work in progress. Includes the number of unfinished items for
products
● over_time: amount of overtime by each team in minutes
● idle_time: The amount of time when the production was interrupted
due to several reasons
● idle_men: The number of workers who were idle due to production
interruption
● no_of_workers: Number of workers in each team
● actual_productivity: The actual % of productivity that was delivered
by the workers. It ranges from 0-1.

The Garment Industry is one of the key examples of the industrial globalization of this
modern era. It is a highly labour-intensive industry with lots of manual processes. Satisfying
the huge global demand for garment products is mostly dependent on the production and
delivery performance of the employees in the garment manufacturing companies. So, it is
highly desirable among the decision makers in the garments industry to track, analyse and
predict the productivity performance of the working teams in their factories.
In particular, they would like to understand what really impacts the productivity of a team.
From a practical point of view, they also know that a productivity score larger than 0.8 is
good enough, while a productivity score lower than 0.8 is not.

1. Build a predictive model for the actual productivity of different teams in different days.
The response in is (0, 1). What can we do to make the regression setting feasible?
2. Build a classification model to understand what teams and in what days generally
have a good enough performance.
Telecom Churn

This dataset contains information about US Telecom customers. Each row represents a
customer. The dataset has the following variables on the columns:

● state: code of the US state of customer residence

● account length: number of months the customer has been with the current telco provider
● area code
● international plan: the customer has an international plan (Yes=1)
● voice mail plan: the customer has a voice-mail plan (Yes=1)
● number vmail messages: number of voice-mail messages
● total day minutes: total minutes of calls during the day
● total day calls: total number of calls during the day
● total day charge: total charge of day charge
● total eve minutes: total minutes of calls during the evening
● total eve calls: total number of calls during the evening
● total eve charge: total charge of evening charge
● total night minutes: total minutes of calls during the night
● total night calls: total number of calls during the night
● total night charge: total charge of night charge
● total intl minutes: total minutes of international calls
● total intl calls: total number of international calls
● total intl charge: total charge of international charge
● number customer service calls: number of calls to customer service
● churn: customer changed telco provider (Yes=1)

Telecom wants to understand what customers are more likely to churn (change provider),
according to the available covariates so that it can target these customers with an ad-hoc
promotional campaign.

1. Build a classification model to help Telecom in targeting the willing-to-churn

customers before it is too late
2. Cluster customers according to their behavior
Korea Real Estate

This dataset contains information about apartment transaction data generated from August
2007 to August 2017 in Daebong district, Daegu city, South Korea. Each row refers to an
apartment. The dataset has the following variables:

● SalePrice: price in US dollar

● YearBuilt
● YearSold
● MonthSold
● Size(sqf): size of apartment in square feet
● Floor: the floor where the apartment is located
● HallwayType
● HeatingType
● AptManageType: how the apartment was managed
● N_Parkinglot: count number of parking spaces on the ground
● TimeToBusStop: measures time takes from apartment to bus stop
● TimeToSubway: measures time takes from apartment to subway station
● N_APT: number of apartments building in an apartment complex
● N_manager: number of people manage apartment facilities (e.g.
security,cleaner,etc.)
● N_elevators: total number of elevators in an apartment complex
● SubwayStation: name of subway station nearby apartment
● N_FacilitiesNearBy(PublicOffices): number of public offices nearby apartment
● N_FacilitiesNearBy(Hospitals): number of hospitals nearby apartment
● N_FacilitiesNearBy(DepStores): number of department stores nearby apartment
● N_FacilitiesNearBy(Malls): number of malls nearby apartment
● N_FacilitiesNearBy(ETC): number of buildings like hotels or special schools nearby
apartment
● N_FacilitiesNearBy(Park): number of parks nearby apartment
● N_SchoolNearBy(Elementary): number of elementary schools nearby apartment
● N_SchoolNearBy(Middle): number of middle schools nearby apartment
● N_SchoolNearBy(High): number of high schools nearby apartment
● N_SchoolNearBy(University): number of universities nearby apartment
● N_FacilitiesInApt: number of facilities for residents like swimming pool, gym,play
ground
● N_FacilitiesNearBy(Total): total number of facilities nearby apartment
● N_SchoolNearBy(Total): total number of schools nearby apartment

A new company wants to enter the real-estate market in Korea, and wants to understand
what kind of houses, and in what areas (facilities, etc.) grant the larger sale prices.

1. Build a predictive model for the sale prices of houses in Korea according to the
available covariates
Car Prices

This dataset contains information about automobiles. Each row represents a different car
model. For each observation we have information about the price and the car features, as
well as an insurance risk score.

● Car_ID: Unique id of each observation (Interger)

● Symboling: Its assigned insurance risk rating, A value of +3 indicates that the auto is
risky, -3 that it is probably pretty safe.(Categorical)
● carCompany: Name of car company (Categorical)
● fueltype: Car fuel type i.e gas or diesel (Categorical)
● aspiration: Aspiration used in a car (Categorical)
● doornumber: Number of doors in a car (Categorical)
● carbody: body of car (Categorical)
● drivewheel: type of drive wheel (Categorical)
● enginelocation : Location of car engine (Categorical)
● wheelbase: Weelbase of car (Numeric)
● carlength: Length of car (Numeric)
● carwidth: Width of car (Numeric)
● carheight: height of car (Numeric)
● curbweight: The weight of a car without occupants or baggage. (Numeric)
● enginetype: Type of engine. (Categorical)
● cylindernumber: cylinder placed in the car (Categorical)
● enginesize: Size of car (Numeric)
● fuelsystem: Fuel system of car (Categorical)
● boreratio: Boreratio of car (Numeric)
● stroke: Stroke or volume inside the engine (Numeric)
● compressionratio: compression ratio of car (Numeric)
● horsepower: Horsepower (Numeric)
● peakrpm: car peak rpm (Numeric)
● citympg: Mileage in city (Numeric)
● highwaympg: Mileage on highway (Numeric)
● price: Price of car (Numeric)

I car producing company has to come up with a new model of car. It wants to target a
specific segment of the market, so it needs a model able to predict the price of the new-
designed car according to its specifics. Furthermore, it is well known that car market is
strongly segmented in different types (van, suv, coupet, etc.). What cluster will the new car
belong?

1. Build a predictive model for the sale prices of cars according to their characteristics
2. Can you get a clusterization of car-type just looking at the available covariates
(excluding model and carbody)?
Affairs

This dataset contains information about extramarital affairs. For each subject the following
features were recorded:

● rate_marriage: How rate marriage, 1 = very poor, 2 = poor, 3 = fair, 4 = good, 5 =

very good
● age: Age
● yrs_married: No. years married. Interval approximations.
● children: No. children
● religious: How religious, 1 = not, 2 = mildly, 3 = fairly, 4 = strongly
● educ: Level of education, 9 = grade school, 12 = high school, 14 = some college,
16= college graduate, 17 = some graduate school, 20 = advanced degree
● occupation: 1 = student, 2 = farming, agriculture; semi-skilled, or unskilled worker;
3 = white-colloar; 4 = teacher, counselor social worker, nurse; artist, writers;
technician, skilled worker, 5 = managerial, administrative, business, 6 = professional
with advanced degree
● occupation_husb: Husband's occupation. Same as occupation.
● affairs: measure of time spent in extramarital affairs

Some possible questions of interest (not exhaustive: these are just some ideas): Is the
amount of time spent in extramarital affairs affected by the other covariates? Does the
presence of children (or other features) tend to affect in some way the probability of being in
a successful marriage?