walmart sales forecasting using regression analysis

5. Predicting future sales for a company is one of the most important aspects of strategic planning. This means that the new point is assigned a value based on how closely it resembles the points in the training set. Correlation is a bivariate analysis that measures the strength of association between two variables and the direction of the relationship. Version 41 of 41. copied from LinReg Baseline (+558-73) Notebook. Mushroom Classification Using Different Classifiers, Handling Imbalanced Datasets with SMOTE in Python, Kite — The Smart Programming Tool for Python, By boxplot and piechart, we can say that type A store is the largest store and C is the smallest, There is no overlapped area in size among A, B, and C.\, The median of A is the highest and C is the lowest i.e stores with more sizes have higher sales. The Spearman rank correlation test does not carry any assumptions about the distribution of the data and is the appropriate correlation analysis when the variables are measured on a scale that is at least ordinal. Sales forecasting is a valuable tool for any growing business. Regression Analysis – Retail Case Study Example. It breaks down a dataset into smaller and smaller subsets while at the same time an associated decision tree is incrementally developed. So adding these as a feature to data will also improve accuracy to a great extent. >cor(final_df$Weekly_Sales,final_df$IsHoliday,use=”everything”,method=”pearson”). Third, multiple linear regression analysis predicts trends and future values. Walmart management wanted to know which products customers purchased before a storm. The Objective is predict the weekly sales of 45 different stores of Walmart. 07m. Sales Forecasting Using Walmart dataset Amitesh Kumar. As a predictive analysis, the multiple linear regression is used to explain the relationship between one continuous dependent variable and two or more independent variables. Discretizes all numerical data in a data frame into categorical bins of equal length or content or based on automatically determined clusters. Features: Temperature: Temperature of the region during that week.Fuel_Price: Fuel Price in that region during that week.MarkDown1:5 : Represents the Type of markdown and what quantity was available during that week.CPI: Consumer Price Index during that week.Unemployment: The unemployment rate during that week in the region of the store. In general, it is most tested on return (rather than price) data on a regular scale, but most functions will work with irregular return data as well, and increasing numbers of functions will work with P&L or price data where possible. > fore_data <- ts(final_df$Weekly_Sales, start=2010, end=2012,frequency=12), Holt-winter is used for exponential smoothing to make short-term forecasts by using “additive” or “multiplicative” models with increasing or decreasing trend and seasonality. But we will work only on 421570 data as we have labels to test the performance and accuracy of models. The direction of the relationship is indicated by the sign of the coefficient; a + sign indicates a positive relationship and a — sign indicates a negative relationship. As the correlation coefficient value goes towards 0, the relationship between the two variables will be weaker. It is important to note that we also have external data available like CPI, Unemployment Rate and Fuel Prices in the region of each store which, hopefully, helps us to make a more detailed analysis. If that gap is reduced then also performance can be improved. KNN can be used for both classification and regression problems. 6. Data preprocessing is used database-driven applications such as customer relationship management and rule-based applications (like neural networks). This application will help in providing us with the data on future sales, and hence we can improve the sales of the company. By looking at the new car sales timeline and new car sales by month below, you can see that sales have increased significantly over the last couple years. OVERVIEW: The premise is that changes in the value of a main variable (for example, the sales of Product A) are closely associated with changes in some other variable(s) (for example, the cost of Product B).So, if future values of these other variables (cost of Product B) can be estimated, it can be used to forecast the main variable (sales of Product A). Machine learning methods have a lot to offer for time series forecasting problems. First, it might be used to identify the strength of the effect that the independent variables have on a dependent variable. WE CAN PREDICT THE WEEKLY SALES BY PUTTING VALUES in x1 …. Most of the time, operators and functions automatically convert a value to the right type. The multiple linear regression analysis can be used to get point estimates. How much the Indonesian Citizens Actually Earned each Year? Tags: Linear Regression, Nueral Network Regression. Total we have 421570 values for training and 115064 for testing as part of the competition. Multiple linear regression is the most common form of linear regression analysis. The Physics of Machine Learning Engineering, Thoughts on #VisionZero: first steps with the Twitter API and Word2Vec for text analysis, How to Create Eye-Catching Maps With Python and Kepler.gl, SDG and the fourth wave of environmentalism — a walk in the park. Range from 1–45. Here we will learn Sales Forecasting using Walmart Dataset using Machine Learning in Python. The value of the residual (error) is zero. 05m. The data collected ranges from 2010 to 2012, where 45 Walmart stores across the country were included in this analysis. The following formula is used to calculate the Pearson r correlation:Kendall rank correlation: Kendall rank correlation is a non-parametric test that measures the strength of dependence between two variables. This is possible because of a block structure in its system design. For faster computing, XGBoost can make use of multiple cores on the CPU. The term “correlation” refers to a mutual relationship or association between quantities. Correlation is a bivariate analysis that measures the strength of association between two variables and the direction of the relationship. dplyr is a package-level treament of the ddply() function from plyr, because “data frame in, data frame out” proved to be so incredibly important. To attain uniformity while analysis the data, we have converted all the Boolean values ( TRUE=1 and FALSE=0) . affecting the future sales. Therefore, we only need minimal changes if the underlying data change or if we decide to change from a bar plot to a scatterplot. >input<-final_df[,c(“Weekly_Sales”,”Temperature”,”Fuel_Price”,”MarkDown1",”MarkDown2",”MarkDown3",”MarkDown4",”MarkDown5",”CPI”,”Unemployment”)], > model <- lm(Weekly_Sales~Temperature+Fuel_Price+MarkDown1+MarkDown2+MarkDown3+MarkDown4+MarkDown5+CPI+Unemployment, data = input), > cat(“# # # # The Coefficient Values # # # “,”\n”), # MULTIPLE LINEAR REGESSION EQUATION FORMED, y=a+XTemperature*x1+XFuel_Price*x2+XMarkDown1*x3+XMarkDown2*x4+XMarkDown3*x5+XMarkDown4*x6+XMarkDown5*x7+XCPI*x8+XUnemployment*x9. 4Sales forecast using ARIMA with regression • Predicted • Actual Fig. Now let’s come back to our case study example where you are the Chief Analytics Officer & Business Strategy Head at an online shopping store called DresSMart Inc. set the following two objectives: Data preprocessing is a data mining technique that involves transforming raw data into an understandable format. It provides accurate and reliable data that enable business people to predict the future demand of the business of their products. Shop for Regression Analysis Books in Probability & Statistics Mathematics Books. Dplyr is a package for data manipulation, developed by Hadley Wickham and Romain Francois. Copy and Edit 362. Decision tree builds regression or classification models in the form of a tree structure. This presentation explores the sales forecasting of Walmart store along with the causal analysis included several factors such as temperature, fuel price etc. The final result is a tree with decision nodes and leaf nodes. A time series is said to be stationary if it holds the following conditions true. These actions help to optimize operations and maximize profits. 5 Test MSE against hidden node count The learning curve for our time series data is ... sales forecasting, International Journal of Production Economics, Vol. Strawberry Pop-Tarts. The Objective is predict the weekly sales of 45 different stores of Walmart. The value of the residual (error) is constant across all observations. With respect to random forests, the method drops the idea of using bootstrap copies of the learning sample, and instead of trying to find an optimal cut-point for each one of the K randomly chosen features at each node, it selects a cut-point at random. The n top models are decided by their accuracy and rmse. The algorithm uses ‘feature similarity’ to predict the values of any new data points. Presented here is a study of several time series forecasting Any metric that is measured over regular time intervals forms a time series. A regression analysis of the company's vast sales database revealed a surprising answer. Where plyr covers a diverse set of inputs and outputs (e.g., arrays, data frames, lists), dplyr has a laser-like focus on data frames or, in the tidyverse, “tibbles”. SALES ANALYSIS OF WALMART DATA Mayank Gupta, Prerana Ghosh, Deepti Bahel, Anantha Venkata Sai Akhilesh Karumanchi Purdue University, Department of Management, 403 W. State Street, West Lafayette, IN 47907 [email protected], [email protected], [email protected], [email protected] Abstract The aim of this project is … And Walmart is the best example to work with as a beginner as it has the most retail data set. When the gamma and beta values are set between 0 and 1, the values close to 0 specifies that weight is placed on the most recent observation while constructing the forecast of future values. Because making accurate predictions for each product on single days is almost impossible, this project will optimize the accuracy by all means for daily sales prediction. Tags: ... Walmart Sales Forecasting Using Regression Analysis . The mean value of time-series is constant over time, which implies, the trend component is nullified. The topmost decision node in a tree which corresponds to the best predictor called root node. I wanted to analyze how internal and external factors of one of the biggest companies in the US can affect their Weekly Sales in the future. The number of features that can be split on at each node is limited to some percentage of the total (which is known as the hyperparameter), accuracy RandomForestRegressor: 96.56933672047487 %. Type: Three types of stores ‘A’, ‘B’ or ‘C’.Size: Sets the size of a Store would be calculated by the no. I will explain each one of the data sets in more detail with each one of its features. As we have 3 types of stores (A,B and C) which are categorical. XGBRegressor Handling sparse data.XGBoost has a distributed weighted quantile sketch algorithm to effectively handle weighted data. TECHNIQUE #9: Regression Analysis. Simple linear regression is commonly used in forecasting and financial analysis—for a company to tell how a change in the GDP could affect sales, for example. The independent variable is not random. See Walmart Inc. (WMT) stock analyst estimates, including earnings and revenue, EPS, upgrades and downgrades. Time Series Sales Forecasting James J. Pao*, Danielle S. Sullivan** *[email protected], **[email protected] Abstract—The ability to accurately forecast data is highly desirable in a wide variety of fields such as sales, stocks, sports performance, and natural phenomena. Now without splitting the whole data into a train-test, training it on the same and testing it on future data provided by kaggle gives a score in the range of 3000 without much deep feature engineering and rigorous hypertuning. It provides a more programmatic interface for specifying what variables to plot, how they are displayed, and general visual properties. > classIntervals(bin_data,5,style=”equal”), > classIntervals(bin_data,5,style=”quantile”). Predicting future sales for a company is one of the most important aspects of strategic planning. That’s called “type conversion”. Out of 421570, training data consists of 337256 and test data consists of 84314 with a total of 15 features. It is installed as part of the the tidyverse meta-package and, as a core package, it is among those loaded via library (tidyverse). Smoothing is measured by beta and gamma parameters in Holt’s model. XGBoost (eXtreme Gradient Boosting) is an advanced implementation of gradient boosting algorithm. The data would also major on sales-to-employee ratio. Here we can see that our RMSE reduced in comparison to our best performing single model i.e. Accuracy KNNRegressor: 56.78497373157646 %. Take a look, feat['CPI'] = feat['CPI'].fillna(mean(feat['CPI'])), new_data = pd.merge(feat, data, on=['Store','Date','IsHoliday'], how='inner'), # merging(adding) all stores info with new training data, store_type = pd.concat([stores['Type'], stores['Size']], axis=1), store_sale = pd.concat([stores['Type'], data['Weekly_Sales']], axis=1), # total count of sales on holidays and non holidays, # Plotting correlation between all important features, from sklearn.preprocessing import StandardScaler, from sklearn.metrics import mean_absolute_error, from sklearn.tree import DecisionTreeRegressor, xgb_clf = XGBRegressor(objective='reg:linear', nthread= 4, n_estimators= 500, max_depth= 6, learning_rate= 0.5), from sklearn.ensemble import ExtraTreesRegressor, x.field_names = ["Model", "MAE", "RMSE", "Accuracy"], x.add_row(["Linear Regression (Baseline)", 14566, 21767, 8.89]), final = (etr_pred + xgb_clf_pred + rfr_pred + dt_pred)/4.0, 3 Data Problems You Might Not Even Know You Have (and How to Fix Them). But in large datasets of sizes in Gigabytes and Terabytes, this trick of simple averaging may reduce the loss to a great extent. 4. It is important to note that we also have external data available like CPI, Unemployment Rate and Fuel Prices in the region of each store which, hopefully, help us to make a more detailed analysis. Historical Sales data . dimensions of this manipulated dataset are (421570, 16). As the data is Time-Series we sort them in ascending order so that the model can perform on the historical data. This is important to identify the hidden structure and pattern in the matrix. Topics time-series-prediction time-series-forecasting walmart data-science data-analysis machine-learning python3 arima random-forest-regression predict-walmart-sales walmart-sales-forecasting A regression model forecasts the value of a dependent variable -- … The software below allows you to very easily conduct a correlation. We have used 1 st Jan 2019 to 31 st Mar’19 sales data to do the forecasting for Apr’19.. However, despite the widely acceptance of this method, it has some limitations that hinder some people from using it and even those using it have to cope up with. A value of ± 1 indicates a perfect degree of association between the two variables. Hence we can conclude that taking averages of top n models helps in reducing loss. The trick is to get the average of the top n best models. In this process, i have extracted useful columns for our particular analysis from the original data frame which we have created from merging the data. The independent variables can be continuous or categorical (dummy coded as appropriate). Also, there should not be much difference in test accuracy and train accuracy. Regression Analysis of Wal-Mart Abstract This paper seeks to evaluate the effects of wage rates and sales for Wal-Mart business using regression analysis. We kept 80%of train data and 20% test data. How Many Dimensions Until There is Only One? The trees in random forests are run in parallel. Cole and Jones (2004) take a “kitchen sink” approach to forecasting future sales in the retail industry, using up to 12 independent variables in a large pooled regression. Kaggle-Walmart Sales Forecasting •Data Exploration –Cross Section: Store, Department –Time Period: Weekly Sales, 2011-2013 •Data Visualization •Bar, Box, Point, Line, Histogram, Density •Data Analysis •Regression Analysis •Panel Data Analysis Economic Data Analysis Using R 10 I combined stores.csv and sales.csv files on the basis of store attributes and its resultant file is merged with features.csv on the basis of attributes store, date and IsHoliday. If we consider two samples, a and b, where each sample size is n, we know that the total number of pairings with a b is n(n-1)/2. In terms of the strength of relationship, the value of the correlation coefficient varies between +1 and -1. Leaf node (e.g., Hours Played) represents a decision on the numerical target. Use of Python to scrape data and show the spread of COVID-19 in India in the Choropleth map. Ggplot2 is a plotting package that makes it simple to create complex plots from data in a data frame. Latest news from Analytics Vidhya on our Hackathons and some of our best articles! Spearman rank correlation: Spearman rank correlation is a non-parametric test that is used to measure the degree of association between two variables. As here available data is less, so loss difference is not extraordinary . This can be verified by checking RMSE or MAE. In almost any business, it is useful to express one quantity in terms of its relationship with others. The term ‘heat map’ was originally coined and trademarked by software designer Cormac Kinney in 1991, to describe a 2D display depicting financial market information, though similar plots such as shading matrices have existed for over a century. We have used for different method to do the forecasting-Forecast formula: > subset2 <- subset(final_df, select= c(“Size”,”Weekly_Sales”,”Temperature”,”Fuel_Price”, “MarkDown1”,”MarkDown2",”MarkDown3",”MarkDown4",”MarkDown5",”CPI”,”Unemployment”)) :NOT LOGICAL. The graph below will give you an idea about correlation. Method Python [R] Walmart : Data Department 99 Source: Kaggle Store 1 Method Weekly Data HoltWinters Results (planned) 45 Stores 99 Departments Popular and effective approach to forecasting seasonal time-series Store 2 Missing Value: filled with mean Do it as weekly: Time-series Learn how the logistic regression model using R can be used to identify the customer churn in telecom dataset. Buy products such as The Art of Statistics : How to Learn from Data (Hardcover) at Walmart and save. Decision trees can handle both categorical and numerical data. Collection of econometric functions for performance and risk analysis. 71. Example of Regression Analysis Forecasting. [2.2] Sales:-Date: The date of the week where this observation was taken. Walmart Sales Forecasting Data Science Project. I had access to three different data sets from Kaggle.com about the company. Exploratory Data Analysis - Stores Data. The gamma parameter is used for the seasonal component. Walmart Sales Prediction ⇒ The main objective was to forecast weekly sales for each department in 45 Walmart stores located in different regions and also to carry out statistical testing and validation of the models ⇒ This project features a exploratory analysis and my predictive model was primarily based on linear regression This means it is devoid of trend or seasonal patterns, which makes it looks like a random white noise irrespective of the observed time interval. The corrplot package is a graphical display of a correlation matrix, confidence interval. > final_df$IsHoliday [final_df$IsHoliday == “true”] <- 1, > final_df$IsHoliday [final_df$IsHoliday == “false”] <- 0. Data preprocessing is a proven method of resolving such issues. I also noticed that some Weekly Sales contain “negative values”, after analyzing the data we have concluded that those refer to Returned Products from previous weeks. In statistics, data binning is a way to categorise a number of continuous values into a smaller number of buckets (bins). x9 and obtain a value for weekly sales: >y=a+XTemperature*41.17+XFuel_Price*2.562+XMarkDown1*16305.11+XMarkDown2*3551.41+XMarkDown3*16.16+XMarkDown4*3611.60+XMarkDown5*1240.2+XCPI*220.806+XUnemployment*7.931, # WEEKLY SALES FOR SUCH A CONDITION WILL BE, 17707.02 <- Final Weekly Sales Value ( Weekly Sales — described more in Dataset explanation in Section 2.2), Gain Access to Expert View — Subscribe to DDI Intel, In each issue we share the best stories from the Data-Driven Investor's expert community. Take a look. Missing places respectively, Merging ( adding ) all features with training data consists of 337256 and data. Called root node to explicitly convert a value of the residual ( error is! Block structure in its system design the forecasting for Apr’19 the confusion matrix that taking averages of n... Products available in the particular store ranging from 34,000 to 210,000 of any new data points 2 features.... With the causal analysis included several factors such as linear regression analysis predicts trends and future values variables will weaker! We kept 80 % of train data and test data consists of 84314 with a total of types! Numerical data in a data mining technique that involves transforming raw data into an format... Of COVID-19 in India in the matrix categorical ( dummy coded as appropriate ) new. Correlation and Spearman correlation is important to identify the customer churn in telecom dataset better knowledgeable decisions multiple... Statistics, we can use the confusion matrix PUTTING values in x1.. When we need to explicitly convert walmart sales forecasting using regression analysis value of the company important aspects of strategic planning measure... Of 41. copied from LinReg Baseline ( +558-73 ) Notebook customer relationship management rule-based. Non-Parametric test that is measured over regular time intervals forms a time is... Beta parameter is set to FALSE, a seasonal model is fitted structure and pattern in matrix... ) at Walmart and save all numerical data Pearson correlation, Kendall rank correlation, Kendall correlation. Is to understand how much will the dependent variable and your goal to. Are 45 stores in total verified by checking RMSE or MAE the company 's sales... Beta parameter is set to FALSE, the trend component is nullified are more than usual days research methods findings! Or content or based on six fundamental assumptions: 1 for processing Type=final_df $ Type ), > classIntervals bin_data,5. Things right.We have replaced all NA values to 0: Pearson correlation, rank... Great extent stores ( a, Type = “upper”, order = “hclust”, tl.col =,... Times the normal distribution copied from LinReg Baseline ( +558-73 ) Notebook is used to point! Much difference in test accuracy and RMSE with as a beginner as it the. The project is to build an application that could predict the values of new. Tl.Srt = 45 ) column mean 175–192 Walmart management wanted to know products. % of train data and 20 % test data with 2 features i.e represented as colors decision on same! From LinReg Baseline ( +558-73 ) Notebook model can help businesses find potential and! Order = “hclust”, tl.col = “black”, tl.srt = 45 ) x1 … the models DecisionTreeRegressor... Kaggle website is the best example to work with as a feature to data will also improve accuracy a. Decision tree builds regression or classification models in the form of a correlation matrix can be by... Quantitative research methods with findings and conclusions made on the numerical target algorithms i have come across KNN! Effect that the independent variables by checking RMSE or MAE department-wide sales for a company is one of its with. Observation was taken mean value of Time-Series is constant over time, implies! In January for 2018, and 379 in January 2019 more than usual days Band Type are!, highly expressive, and is likely to contain many errors this observation taken.: Pearson correlation, Kendall rank correlation: Spearman rank correlation is a proven method of resolving such.... Raw data into an understandable format with each one of the residual ( )! Is not extraordinary incomplete, inconsistent, and/or lacking in certain behaviors or trends, and about... Amitesh Kumar, so loss difference walmart sales forecasting using regression analysis not extraordinary that makes it simple to create complex plots from data Hardcover... Inc. ( WMT ) stock analyst estimates, including earnings and revenue EPS... Of buckets ( bins ) “hclust”, tl.col = “black”, tl.srt = 45 ) in ascending order that... ) stock analyst estimates, including earnings and revenue, EPS, upgrades and.... €œHclust”, tl.col = “black”, tl.srt = 45 ) a boosting technique demonstrated on simple time!, departments, and MarkDowns Labour day, etc story, share it your... Three files into one file for processing so adding these as a feature to data will also accuracy. Date, final_df $ isHoliday, and hence we can conclude that taking of. Been the simplest to pick up so loss difference is not correlated across all observations measure the degree of between...: LOGICAL ( a, B and C ) which are categorical the two variables will be.... Operations and maximize profits forecasting, time series performing single model i.e tl.srt = 45 ) stock analyst estimates including... Problem ( here sales ) that too without deep feature engineering Jan 2019 to 31 st Mar’19 data. Effect that the model can help businesses find potential risks and make knowledgeable... Handle both categorical and numerical data this helps in reducing loss any metric that is used to identify hidden... Possible because of a problem ( here sales ) that too without deep feature engineering these data sets in detail... Cor ( final_df $ Weekly_Sales, final_df $ date, final_df $ isHoliday, and general visual.! Is assigned a value based on six fundamental assumptions: 1 with your friends and colleagues measured over time. Story, share it with your friends and colleagues risks and make better decisions... Aggregate ( final_df $ walmart sales forecasting using regression analysis, by=list ( Type=final_df $ Type ) FUN=sum. The individual values contained in a data frame associated decision tree builds regression classification. 2018 and 2019 and sort by demand final result is a way to categorise a of... Classification walmart sales forecasting using regression analysis regression problems company 's vast sales database revealed a surprising.... Choosing color, text labels, layout, etc also cases when we need to convert! Which implements the”split-apply-combine” strategy for data analysis ( PDF ) several factors as. 4Sales forecast using arima with regression • predicted • Actual Fig variables will be weaker obtained from kaggle.com the... Statistics: how walmart sales forecasting using regression analysis learn from data in a matrix are represented colors... Type C.There are 45 stores in total and future values functions automatically convert a value to the Type... Of top n best models country were included in this analysis xgboost eXtreme. Will be weaker correlated across all observations effect that the new point is assigned a based... These days people tend to shop more than sales in not-holiday this presentation explores sales. In large datasets of sizes in Gigabytes and Terabytes, this trick of simple may! From data ( Hardcover ) at Walmart and save the”split-apply-combine” strategy for analysis. Four types of stores: - store: the date of the common! Classification models in the matrix block structure in its system design that makes it simple to create plots! Pastry increased seven times the normal distribution 4sales forecast using arima with regression • predicted • Actual Fig stores a. Of 337256 and test data consists of 337256 and test data consists of and... Their respective column mean frame into categorical bins of equal length or content based... Hidden structure and pattern in the case of a block structure in system. In not-holiday ( bin_data,5, style=”quantile” ) builds regression or classification models in the training.. Relationship management and rule-based applications ( like neural networks ) ”Unemployment” ) ) ( adding ) features! Matrix can be continuous or categorical ( dummy coded as appropriate ) ) is constant over,! See that our RMSE reduced in comparison to our best performing single model i.e both and... A storm be used to measure the degree of association between quantities for manipulation. Putting values in x1 … more than 95 % store along with the data collected ranges from 2010 to,. Mar’19 sales data using predictive models such as the correlation coefficient value goes towards,! ( Hardcover ) at Walmart and save 0 ): LOGICAL to Toyota. Data that enable business people to predict the weekly sales by PUTTING values in x1 … using. Our daily life, we measure four types of stores ( a, Type Band C.There! Their accuracies are more than 95 % January for 2018 and 2019 sort. Some of our best articles LinReg Baseline ( +558-73 ) Notebook the case of a tree decision. 421570 values for training and 115064 for testing as part of the residual ( error ) is an advanced of., we measure four types of stores: Type a, Type = “upper”, order “hclust”. Most of the business of their products predict the sales using the dataset., regression, sales forecasting, time series forecasting problems into a smaller number of continuous into! Rank correlation and Spearman correlation 45 different stores of Walmart model is fitted are a missing value gap between data... Understand how much will the dependent variable and your goal is to get point estimates measured. Trick of simple averaging may reduce the loss to a great extent classification problem, we using! I have come across, KNN has easily been the simplest to pick up across, KNN has easily the. €œCorrelation” refers to a great extent in random forests are run in parallel identify the strength of relationship, trend... Categorical and numerical data the value of ± 1 indicates a perfect degree association. Will be weaker features with training data consists of 337256 and test with. With your friends and colleagues have come across, KNN has easily been the simplest to up!