Data in social and behavioral sciences are routinely collected using questionnaires, and each domain of interest is tapped by multiple indicators. Structural equation modeling (SEM) is one of the most widely used methods to analyze such data. However, conventional methods for SEM face difficulty when the number of variables (p) is large even when the sample size (N) is also rather large. This article addresses the issue of model inference with the likelihood ratio statistic Tml. Using the method of empirical modeling, mean-and-variance corrected statistics for SEM with many variables are developed. Results show that the new statistics not only perform much better than Tml but also are substantial improvements over other corrections to Tml. When combined with a robust transformation, the new statistics also perform well with non-normally distributed data.

China Industrial Enterprise: Number of Employee: Average data was reported at 73,200.000 Person th in Oct 2025. This records a decrease from the previous number of 73,339.000 Person th for Sep 2025. China Industrial Enterprise: Number of Employee: Average data is updated monthly, averaging 91,201.000 Person th from Dec 1992 (Median) to Oct 2025, with 193 observations. The data reached an all-time high of 99,772.100 Person th in Dec 2014 and a record low of 54,408.390 Person th in Dec 2001. China Industrial Enterprise: Number of Employee: Average data remains active status in CEIC and is reported by National Bureau of Statistics. The data is categorized under China Premium Database’s Industrial Sector – Table CN.BF: Industrial Financial Data.

This paper explores a unique dataset of all the SET ratings provided by students of one university in Poland at the end of the winter semester of the 2020/2021 academic year. The SET questionnaire used by this university is presented in Appendix 1. The dataset is unique for several reasons. It covers all SET surveys filled by students in all fields and levels of study offered by the university. In the period analysed, the university was entirely in the online regime amid the Covid-19 pandemic. While the expected learning outcomes formally have not been changed, the online mode of study could have affected the grading policy and could have implications for some of the studied SET biases. This Covid-19 effect is captured by econometric models and discussed in the paper. The average SET scores were matched with the characteristics of the teacher for degree, seniority, gender, and SET scores in the past six semesters; the course characteristics for time of day, day of the week, course type, course breadth, class duration, and class size; the attributes of the SET survey responses as the percentage of students providing SET feedback; and the grades of the course for the mean, standard deviation, and percentage failed. Data on course grades are also available for the previous six semesters. This rich dataset allows many of the biases reported in the literature to be tested for and new hypotheses to be formulated, as presented in the introduction section. The unit of observation or the single row in the data set is identified by three parameters: teacher unique id (j), course unique id (k) and the question number in the SET questionnaire (n ϵ {1, 2, 3, 4, 5, 6, 7, 8, 9} ). It means that for each pair (j,k), we have nine rows, one for each SET survey question, or sometimes less when students did not answer one of the SET questions at all. For example, the dependent variable SET_score_avg(j,k,n) for the triplet (j=Calculus, k=John Smith, n=2) is calculated as the average of all Likert-scale answers to question nr 2 in the SET survey distributed to all students that took the Calculus course taught by John Smith. The data set has 8,015 such observations or rows. The full list of variables or columns in the data set included in the analysis is presented in the attached filesection. Their description refers to the triplet (teacher id = j, course id = k, question number = n). When the last value of the triplet (n) is dropped, it means that the variable takes the same values for all n ϵ {1, 2, 3, 4, 5, 6, 7, 8, 9}.Two attachments:- word file with variables description- Rdata file with the data set (for R language).Appendix 1. Appendix 1. The SET questionnaire was used for this paper. Evaluation survey of the teaching staff of [university name] Please, complete the following evaluation form, which aims to assess the lecturer’s performance. Only one answer should be indicated for each question. The answers are coded in the following way: 5- I strongly agree; 4- I agree; 3- Neutral; 2- I don’t agree; 1- I strongly don’t agree. Questions 1 2 3 4 5 I learnt a lot during the course. ○ ○ ○ ○ ○ I think that the knowledge acquired during the course is very useful. ○ ○ ○ ○ ○ The professor used activities to make the class more engaging. ○ ○ ○ ○ ○ If it was possible, I would enroll for the course conducted by this lecturer again. ○ ○ ○ ○ ○ The classes started on time. ○ ○ ○ ○ ○ The lecturer always used time efficiently. ○ ○ ○ ○ ○ The lecturer delivered the class content in an understandable and efficient way. ○ ○ ○ ○ ○ The lecturer was available when we had doubts. ○ ○ ○ ○ ○ The lecturer treated all students equally regardless of their race, background and ethnicity. ○ ○

Supporting tables and figures. Table S1. The impact of different effect sizes on gene selection strategies when the sample size is fixed and relatively small. Mean (STD) of true positives computed from SIMU1 with 20 repetitions are reported. Sample size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S2. The impact of different effect sizes on gene selection strategies when the sample size is fixed and relatively small. Mean (STD) of false positives computed from SIMU1 with 20 repetitions are reported. Sample size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S3. The impact of different sample sizes on gene selection strategies when the effect size is fixed and relatively small. Mean (STD) of true positives computed from SIMU2 with 20 repetitions are reported. Effect size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S4. The impact of different sample sizes on gene selection strategies when the effect size is fixed and relatively small. Mean (STD) of false positives computed from SIMU2 with 20 repetitions are reported. Effect size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S5. The impact of different sample sizes on gene selection strategies when the effect size is fixed and relatively large. Mean (STD) of true positives computed from SIMU2 with 20 repetitions are reported. Effect size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S6. The impact of different sample sizes on gene selection strategies when the effect size is fixed and relatively large. Mean (STD) of false positives computed from SIMU2 with 20 repetitions are reported. Effect size: . Total number of genes: 1000. Number of differentially expressed genes: 100. Number of permutations for Nstat: 10000. The significance threshold: 0.05. Table S7. The impact of different sample sizes on gene selection strategies with simulation based on biological data. Mean (STD) of true positives computed from SIMU-BIO with 20 repetitions are reported. Total number of genes: 9005. Number of permutations for Nstat: 100000. The significance threshold: 0.05. Table S8. The impact of different sample sizes on gene selection strategies with simulation based on biological data. Mean (STD) of false positives computed from SIMU-BIO with 20 repetitions are reported. Total number of genes: 9005. Number of permutations for Nstat: 100000. The significance threshold: 0.05. Table S9. The numbers of differentially expressed genes detected by different selection strategies. Total number of genes: 9005. Number of permutations for Nstat: 100000. The significance threshold: 0.05. Figure S1. Histogram of pairwise Pearson correlation coefficients between genes computed from HYPERDIP without normalization. Number of genes: 9005. Number of arrays: 88. (PDF)

Bangladesh number dataset provides contact information from trusted sources. We only collect phone numbers from reliable sources and define this information. To ensure transparency, we also provide the source URL to show where the information was collected from. In addition, we offer 24/7 support. If you have a question or need help, we’re always here. However, we care about accuracy, so we carefully collect the Bangladesh number dataset from trusted sources. You may rely on this data for business or personal use. With customer support, you’ll never have to wait when you need help or more information. We use opt-in data to respect privacy. This way, we contact only people who want to hear from you. Bangladesh phone data gives you access to contacts in Bangladesh. Here you can filter information by gender, age, and relationship status. This makes it easy to find exactly the people you want to connect with. We define this data by ensuring it follows all GDPR rules to keep it safe and legal. Our system works hard to remove any invalid data so you get only accurate and valid numbers. List to Data is a helpful website for finding important phone numbers quickly. Also, our Bangladesh phone data is suitable for doing business targeting specific groups. You can easily filter your list to focus on specific types of customers. Since we remove invalid data regularly, you don’t have to deal with old or useless numbers. We assure you that all data follows strict GDPR rules, so you can use it without any problems. Bangladesh phone number list is a collection of phone numbers from people in Bangladesh. We define this list by providing 100% correct and valid phone numbers that are ready to use. Also, we offer a replacement guarantee if you ever receive an invalid number. This means you will always have accurate data. We collect phone numbers that we provide based on customer’s permission. Moreover, we work hard to provide the best Bangladesh phone number list for businesses and personal use. We gather data correctly, so you won’t have to worry about getting outdated or incorrect information. Our replacement guarantee means you’ll always have valid numbers, so you can relax and feel confident.

Thailand Average Monthly Revenue: Per Mobile Phone Number: Prepaid data was reported at 151.000 THB in Sep 2019. This records a decrease from the previous number of 152.000 THB for Jun 2019. Thailand Average Monthly Revenue: Per Mobile Phone Number: Prepaid data is updated quarterly, averaging 152.000 THB from Mar 2014 (Median) to Sep 2019, with 23 observations. The data reached an all-time high of 165.000 THB in Mar 2016 and a record low of 134.000 THB in Sep 2014. Thailand Average Monthly Revenue: Per Mobile Phone Number: Prepaid data remains active status in CEIC and is reported by Office of The National Broadcasting and Telecommunications Commission. The data is categorized under Global Database’s Thailand – Table TH.TB006: Telecommunication Statistics: Office of The National Broadcasting and Telecommunications Commission .

Per-Title Encoding

Nowadays video streaming plays an important role in our daily lives. To achieve the best video quality, some important attributes should be taken into consideration. For example internet bandwidth is one those. Some companies like Netflix, Amazon and YouTube have tried to provide the best video quality according to different internet bandwidths to satisfy their user’s demands, therefore these companies have offered some solutions to reach this goal.

one of the main approaches which is suitable for most of the videos having a diverse set of pixel complexities, is the fixed bitrate ladder approach. This approach came with a set of problems such as video artefacts in high-complexity videos or waste of bandwidth, in low-complex videos.

One optimal quality metric approach developed by Netflix is VMAF (Video Multi-Method Assessment Fusion).Here we have three machine learning methods Linear Regression, Support Vector Regression & Random Forest to predict VMAF to find the optimal video quality for different videos based on their content.

Data

Data is one csv file which contains four sections Source,Video Characteristic,Encoding Setting and Target Value

• Video_data_set.csv - A collected data from videos

• Source Viedo

  • s_video_id - Identifer for the original video (numerical)
  • s_width - Resolution of the raw video (numerical)
  • s_height - Resolution of the raw video (numerical)
  • s_storage_size - the total size of the source video without audio tracks (numerical)
  • s_duration - Length of the source video (numerical)
  • s_scan_type - "progressive" or "interlaced"(categorical)

----------------------------------------------------------------------------------

• Video Characteristic

  • c_content_category - A label indicating the category of the video with the highest probability(categorical)
  • c_scene_change_ffmp_ratio30 - Indicates how many scene changes appear per minute on average in the video for a given probability 30%(numerical)
  • c_scene_change_ffmp_ratio60 - Indicates how many scene changes appear per minute on average in the video for a given probability 60%(numerical)
  • c_scene_change_ffmp_ratio90 - Indicates how many scene changes appear per minute on average in the video for a given probability 90%(numerical)
  • c_scene_change_py_thresh30 - Indicates how many scene changes appear throughout entire clip with threshold 30(numerical)
  • c_scene_change_py_thresh50 - Indicates how many scene changes appear throughout entire clip with threshold 50(numerical)
  • c_si - The spatial perceptual information (SI) based on the Sobel filter averaged over the whole video(numerical)
  • c_ti - The temporal perceptual information (TI) based upon the motion difference feature averaged over the whole video(numerical)
  • c_colorhistogram_mean_dark - color values between [0-63] are grouped in this block (dark)population mean of RGB color values normalised - divide by pixel count & divide by channel count mean of mean over all frames of a video(numerical)
  • c_colorhistogram_mean_medium_dark - color values between 64-127
  • c_colorhistogram_mean_medium_bright - color values between 128-195
  • c_colorhistogram_mean_bright - color values between 196-255
  • c_colorhistogram_std_dev_dark - standard deviation of c_colorhistogram_mean_medium_dark within each frame mean of all frames of a video(numerical)
  • c_colorhistogram_std_dev_medium_dark - standard deviation of c_colorhistogram_mean_medium_bright within each frame mean of all frames of a video(numerical)
  • c_colorhistogram_std_dev_medium_bright - standard deviation of c_colorhistogram_mean_medium_bright within each frame mean of all frames of a video(numerical)
  • c_colorhistogram_std_dev_bright - standard deviation of c_colorhistogram_mean_bright within each frame mean of all frames of a video(numerical)
  • c_colorhistogram_temporal_mean_std_dev_dark - temporal standard deviation of mean of c_colorhistogram_mean_dark(numerical)
  • c_colorhistogram_temporal_mean_std_dev_medium_dark - temporal standard deviation of mean of c_colorhistogram_mean_medium_dark(numerical)
  • c_colorhistogram_temporal_mean_std_dev_medium_bright - temporal standard deviation of mean of c_colorhistogram_mean_medium_bright(numerical)
  • c_colorhistogram_temporal_mean_std_dev_bright - temporal standard deviation of mean of c_colorhistogram_mean_bright(numerical)

• Encoding Setting

  • e_crf - Constant Rate Factor for this encoding(numerical)
  • e_width - Target Resolution of the encoded video(numerical)
  • e_height - Target Resolution of the encoded video(numerical)
  • e_aspect_ratio - Aspect ratio of the video(numerical)
  • e_pixel_aspect_ratio - Aspect ratio of the pixels. Usu...

Uruguay UY: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data was reported at 2.900 NA in 2017. This records an increase from the previous number of 2.600 NA for 2010. Uruguay UY: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data is updated yearly, averaging 2.600 NA from Dec 2006 (Median) to 2017, with 3 observations. The data reached an all-time high of 2.900 NA in 2017 and a record low of 2.200 NA in 2006. Uruguay UY: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data remains active status in CEIC and is reported by World Bank. The data is categorized under Global Database’s Uruguay – Table UY.World Bank: Company Statistics. Average number of visits or required meetings with tax officials during the year. The value represents the average number of visits for all firms which reported being visited or required to meet with tax officials (please see indicator IC.FRM.METG.ZS).; ; World Bank, Enterprise Surveys (http://www.enterprisesurveys.org/).; Unweighted average;

The data describes the total number of casualties of each war in the time span 768CE till 2019. This data is collected from the website www.necrometrics.com where for each war in the time span a number of sources are given each source providing an estimate of the number of casualties in the war. Along with the minimum and maximum estimates of the total number of casualties, respectively the median and mean values of these sources are collected and have been put into an excel file. The data can be used to analyze wars and the severity of wars in the past.

Nicaragua NI: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data was reported at 3.700 NA in 2016. This records an increase from the previous number of 2.700 NA for 2010. Nicaragua NI: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data is updated yearly, averaging 2.700 NA from Dec 2006 (Median) to 2016, with 3 observations. The data reached an all-time high of 3.700 NA in 2016 and a record low of 2.500 NA in 2006. Nicaragua NI: Number of Visits or Required Meetings with Tax Officials: Average for Affected Firms data remains active status in CEIC and is reported by World Bank. The data is categorized under Global Database’s Nicaragua – Table NI.World Bank.WDI: Company Statistics. Average number of visits or required meetings with tax officials during the year. The value represents the average number of visits for all firms which reported being visited or required to meet with tax officials (please see indicator IC.FRM.METG.ZS).; ; World Bank, Enterprise Surveys (http://www.enterprisesurveys.org/).; Unweighted average;

China Paper Product: Number of Employee: Average data was reported at 642.599 Person th in Dec 2013. This records a decrease from the previous number of 703.299 Person th for Dec 2012. China Paper Product: Number of Employee: Average data is updated monthly, averaging 620.800 Person th from Dec 1998 (Median) to Dec 2013, with 69 observations. The data reached an all-time high of 773.900 Person th in Dec 2010 and a record low of 308.200 Person th in Dec 2000. China Paper Product: Number of Employee: Average data remains active status in CEIC and is reported by National Bureau of Statistics. The data is categorized under China Premium Database’s Industrial Sector – Table CN.BHJ: Paper Making: Paper Product.

Japan Average Number of Nights: Total data was reported at 5.838 Night in Jun 2018. This records an increase from the previous number of 5.635 Night for Mar 2018. Japan Average Number of Nights: Total data is updated quarterly, averaging 5.932 Night from Mar 2014 (Median) to Jun 2018, with 18 observations. The data reached an all-time high of 6.247 Night in Dec 2014 and a record low of 5.437 Night in Mar 2015. Japan Average Number of Nights: Total data remains active status in CEIC and is reported by Ministry of Land, Infrastructure, Transport and Tourism. The data is categorized under Global Database’s Japan – Table JP.Q030: Tourism and Leisure: Average Number of Nights Stay by Nationality.

Synthetic data for assessing and comparing local post-hoc explanation of detected process shift

DOI

10.5281/zenodo.15000635

Synthetic dataset contains data used in experiment described in article submitted to Computers in Industry journal entitled

Assessing and Comparing Local Post-hoc Explanation for Shift Detection in Process Monitoring.

The citation will be updated immediately after the article will be accepted.

Particular data.mat files are stored in a subfolder structure, which clearly assigns the particular file to

on of the tested cases.

For example, data for experiments with normally distributed data, known number of shifted variables and 5 variables are stored in path ormal\known_number\5_vars\rho0.1.

The meaning of particular folders is explained here:

normal - all variables are normally distributed

not-normal - copula based multivariate distribution based on normal and gamma marginal distributions and defined correlation

known_number - known number of shifted variables (methods used this information, which is not available in real world)

unknown_number - unknown number of shifted variables, realistic case

2_vars - data with 2 variables (n=2)

...

10_vars - data with 10 variables (n=2)

rho0.1 - correlation among all variables is 0.1

...

rho0.9 - correlation among all variables is 0.9

Each data.mat file contains the following variables:

LIME_res nval x n results of LIME explanation

MYT_res nval x n results of MYT explanation

NN_res nval x n results of ANN explanation

X p x 11000 Unshifted data

S n x n sigma matrix (covariance matrix) for the unshifted data

mu 1xn mean parameter for the unshifted data

n 1x1 number of variables (dimensionality)

trn_set n x ntrn x 2 train set for ANN explainer,

         trn_set(:,:,1) are values of variables from shifted process

         trn_set(:,:,2) labels denoting which variables are shifted 

         trn_set(i,j,2) is 1 if ith variable of jth sample trn_set(:,j,1) is shifted

val_set n x 95 x 2 validation set used for testing and generating LIME_res, MYT_res and NN_res

Magnetohydrodynamic (MHD) turbulence is revealed to have scaling anisotropy based on a local mean magnetic field, which is usually calculated through the average of the magnetic field vectors at two points. However, the thus reconstructed parallel spectrum can be contaminated by intermittent events and nonparallel fluctuations. In this work, we simulate the driven compressible MHD turbulence to investigate the influence of the stationary local mean field on the scaling anisotropy. By examining whether the local parallel samplings are along the magnetic field directions, we confirm that the stationary local mean field can be fulfilled by the criterion ϕ < 10◦ with ϕ being the angle between the two local mean magnetic fields after cutting the sampled magnetic field intervals into two halves. The three-dimensional (3D) numerical data also show that the samplings with the stationary local mean fields successfully prevent intermittency events and nonparallel fluctuations from mixing into the measured parallel fluctuations. With the true parallel samplings achieved, the local parallel spectra of the magnetic field and the velocity are found to be closer to -5/3 scaling than -2 scaling. The requirement of the stationary local mean fields has nearly no effects on the perpendicular scaling, which has a about -5/3 power-law index. Among the removed intermittent events from the local parallel samplings, one third of the events are identified as TDs, one third as RDs, and the rest as EDs. The samplings with the nonstationary local mean fields display large separations between the field directions and the parallel directions, about two times of those with the stationary local mean fields according to statistical analyses. These results have important implications for interpreting the nonlinear interactions of MHD turbulence.

This large longitudinal study is the result of professor Tatu Vanhanen's long-term research on democratization and power resources. International scientific community knows this data also by the name "Vanhanen's Index of Power Resources". The data have been collected from several written sources and have been published as appendices of five different books. The books are listed in the section Data sources below. The original sources of the numerical data published in these books have been collected to a separate document containing background information.

Vanhanen divides the variables of his dataset into two main groups. The first group consists of Measures of Democracy and includes three variables. The second group is called Measures of Resource Distribution.

The variables in the first group (Measures of Democracy) are Competition, Participation and Index of Democratization. The value of Competition is calculated by subtracting the percentage of votes/seats gained by the largest political party in parliamentary elections and/or in presidential (executive) elections from 100%. The Participation variable is an aggregate of the turnout in elections (percentage of the total population who voted in the same election) and the number of referendums. Each national referendum raises the value of Participation by five percentage points and each state referendum by one percentage point for the year of the referendum. The upper limit for both variables is 70%. Index of Democratization is derived by first multiplying the above mentioned variables Competition and Participation and then dividing this product by 100.

Six variables are used to measure resource distribution: 1) Urban Population (%) (as a percentage of total population). 2) Non-Agricultural Population (%) (derived by subtracting the percentage of agricultural population from 100%). 3) Number of students: the variable denotes how many students there are in universities and other higher education institutions per 100.000 inhabitants of the country. Two ways are used to calculate the percentage of Students (%): before the year 1988 the value 1000 of the variable Number of students is equivalent to 100% and between the years 1988-1998 the value 5000 of the same variable is equivalent to 100%. 4) Literates (%) (as a percentage of adult population). 5) Family Farms Are (%) (as a percentage of total cultivated area or of total area of holdings). 6) Degree of Decentralization of Non-Agricultural Economic Resources. This variable has been calculated from the 1970s.

Three new variables have been derived from the above mentioned six variables. 1) Index of Occupational Diversification is derived by calculating the arithmetic mean of Urban Population and Non-Agricultural Population. 2) Index of Knowledge Distribution is derived by calculating the arithmetic mean of Students and Literates. 3) Index of Distribution of Economic Power Resources is derived by first multiplying the value of Family Farm Area with the percentage of agricultural population. Then the value of Degree of Decentralization of Non-Agricultural Economic Resources is multiplied with the percentage of Non-Agricultural Population. After this these two products are simply added up.

Finally two new variables have derived from the above mentioned variables. First derived variable is Index of Power Resources, calculated by multiplying the values of Index of Occupational Diversification, Index of Knowledge Distribution and Index of the Distribution of Economic Power Resources and then dividing the product by 10 000. The second derived variable Mean is the arithmetic mean of the five (from the 1970s six) explanatory variables. This differs from Index of Power Resources in that a low value of any single variable does not reduce the value of Mean to any great extent.

Russia whatsapp number list is a collection of whatsapp numbers from people living in Russia. This list is very useful for businesses and organizations that want to reach out to these individuals. The numbers in this list are 100% correct and valid. This means that every number works, so businesses can call confidently. If any number does not work, you receive a replacement guarantee. This guarantee means the company will give you a new number at no extra cost. This way, you always have reliable contacts. At List to Data, we help you find important whatsapp numbers easily and quickly. Russia whatsapp phone number data is a valuable database that allows businesses to filter information based on specific needs. This means you can filter the data by gender, age, and relationship status. For example, if a business wants to reach younger people, it can easily find numbers for that age group. This ability to filter information makes communication more effective. You can focus on the audience that matters most to you. Additionally, the database follows GDPR rules, which protect people’s privacy. Following these rules ensures that all data usage is legal and ethical. List to Data helps you find whatsapp numbers for your busines

What does the data show?

This data shows the monthly averages of maximum surface temperature (°C) for 2040-2069 using a combination of the CRU TS (v. 4.06) and UKCP18 global RCP2.6 datasets. The RCP2.6 scenario is an aggressive mitigation scenario where greenhouse gas emissions are strongly reduced.

The data combines a baseline (1981-2010) value from CRU TS (v. 4.06) with an anomaly from UKCP18 global. Where the anomaly is the change in temperature at 2040-2069 relative to 1981-2010.

The data is provided on the WGS84 grid which measures approximately 60km x 60km (latitude x longitude) at the equator.

Limitations of the data

We recommend the use of multiple grid cells or an average of grid cells around a point of interest to help users get a sense of the variability in the area. This will provide a more robust set of values for informing decisions based on the data.

What are the naming conventions and how do I explore the data?

This data contains a field for each month’s average over the period. They are named 'tmax' (temperature maximum), the month and ‘upper’ ‘median’ or ‘lower’. E.g. ‘tmax Mar Lower’ is the average of the daily minimum temperatures in March throughout 2040-2069, in the second lowest ensemble member.

To understand how to explore the data, see this page: https://storymaps.arcgis.com/stories/457e7a2bc73e40b089fac0e47c63a578

Please note, if viewing in ArcGIS Map Viewer, the map will default to ‘tmax Jan Median’ values.

What do the ‘median’, ‘upper’, and ‘lower’ values mean?

Climate models are numerical representations of the climate system. To capture uncertainty in projections for the future, an ensemble, or group, of climate models are run. Each ensemble member has slightly different starting conditions or model set-ups. Considering all of the model outcomes gives users a range of plausible conditions which could occur in the future.

To select which ensemble members to use, the monthly averages of maximum surface temperature for the period 2040-2069 were calculated for each ensemble member and they were then ranked in order from lowest to highest for each location.

The ‘lower’ fields are the second lowest ranked ensemble member. The ‘upper’ fields are the second highest ranked ensemble member. The ‘median’ field is the central value of the ensemble.

This gives a median value, and a spread of the ensemble members indicating the range of possible outcomes in the projections. This spread of outputs can be used to infer the uncertainty in the projections. The larger the difference between the lower and upper fields, the greater the uncertainty.

Data source

CRU TS v. 4.06 - (downloaded 12/07/22)

UKCP18 v.20200110 (downloaded 17/08/22)

Useful links

Further information on CRU TS Further information on the UK Climate Projections (UKCP) Further information on understanding climate data within the Met Office Climate Data Portal

Description

The data is related with direct marketing campaigns of a Portuguese banking institution. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required, in order to access if the product (bank term deposit) would be ('yes') or not ('no') subscribed.

Input variables:

Bank client data 1 - age (numeric) 2 - job : type of job (categorical: 'admin.','blue-collar','entrepreneur','housemaid','management','retired','self-employed','services','student','technician','unemployed','unknown') 3 - marital : marital status (categorical: 'divorced','married','single','unknown'; note: 'divorced' means divorced or widowed) 4 - education (categorical: 'basic.4y','basic.6y','basic.9y','high.school','illiterate','professional.course','university.degree','unknown') 5 - default: has credit in default? (categorical: 'no','yes','unknown') 6 - housing: has housing loan? (categorical: 'no','yes','unknown') 7 - loan: has personal loan? (categorical: 'no','yes','unknown')

related with the last contact of the current campaign:

8 - contact: contact communication type (categorical: 'cellular','telephone') 9 - month: last contact month of year (categorical: 'jan', 'feb', 'mar', ..., 'nov', 'dec') 10 - day_of_week: last contact day of the week (categorical: 'mon','tue','wed','thu','fri') 11 - duration: last contact duration, in seconds (numeric). Important note: this attribute highly affects the output target (e.g., if duration=0 then y='no'). Yet, the duration is not known before a call is performed. Also, after the end of the call y is obviously known. Thus, this input should only be included for benchmark purposes and should be discarded if the intention is to have a realistic predictive model. Other attributes 12 - campaign: number of contacts performed during this campaign and for this client (numeric, includes last contact) 13 - pdays: number of days that passed by after the client was last contacted from a previous campaign (numeric; 999 means client was not previously contacted) 14 - previous: number of contacts performed before this campaign and for this client (numeric) 15 - poutcome: outcome of the previous marketing campaign (categorical: 'failure','nonexistent','success')

social and economic context attributes

16 - emp.var.rate: employment variation rate - quarterly indicator (numeric) 17 - cons.price.idx: consumer price index - monthly indicator (numeric) 18 - cons.conf.idx: consumer confidence index - monthly indicator (numeric) 19 - euribor3m: euribor 3 month rate - daily indicator (numeric) 20 - nr.employed: number of employees - quarterly indicator (numeric)

Output variable 21 - y - has the client subscribed a term deposit? (binary: 'yes','no')