Sports-Wear-Group

Objective And Problem Statement

Focus Of Analytics Project :

• marketing team will be interest to see if they can increase their efficiency
• the management team want to know if using advanced analytics may increase sales in general.

Attributes:

• country: Country name
• article: 6 digit article number, as unique identifier of an article
• sales: total number of units sold in respective retail week
• regular_price : recommended retail price of the article
• current_price : current selling price (weighted average over the week)
• ratio : price ratio as current_price/regular_price, such that price discount is 1-ratio
• retailweek : start date of the retailweek
• promo1 : indicator for media advertisement, taking 1 in weeks of activation and 0 otherwise
• promo2 : indicator for store events, taking 1 in weeks with events and 0 otherwise
• customer_id : customer unique identifier, one id per customer
• article : 6 digit article number, as unique identifier of an article
• productgroup : product group the article belongs to
• category : product category the article belongs to
• cost : total costs of the article (assumed to be fixed over time)
• style : description of article design
• sizes : size range in which article is available
• gender : gender of target consumer of the article
• rgb_*_main_color: intensity of the red (r), green (g), and blue (b) primaries of the article‘s main color, taking values [0,250]
• rgb_*_sec_color: intensity of the red (r), green (g), and blue (b) primaries of the article‘s secondary color, taking values [0,250]
• label: advertisement result after offering/sending/presenting the offer to the customer. 0 means the customer did not buy and 1 means the costumer did buy.

Description

• article 477 distinct values High cardinality
• retailweek 123 distinct Weeks High cardinality
• regular_price is highly overall correlated with current_price High correlation
• current_price is highly overall correlated with regular_price High correlation
• productgroup is highly overall correlated with cost High correlation
• category is highly overall correlated with cost High correlation
• style is highly overall correlated with cost High correlation
• sizes is highly overall correlated with cost High correlation
• gender is highly overall correlated with cost High correlation
• closest_colors_main is highly overall correlated with cost High correlation
• closest_colors_sec is highly overall correlated with cost High correlation
• promo1 is highly imbalanced (66.5%) Imbalance
• promo2 is highly imbalanced (95.5%) Imbalance
• sizes is highly imbalanced (53.1%) Imbalance
• article.1 is uniformly distributed Uniform