(in Polish) Advanced statistics

Basics in quantitative data analysis. Knowledge of basic statistical terms.

compulsory course

Contact hours with teacher:

- participation in lectures - 15 hrs

- consultations- 15 hrs

Self-study hours:

- preparation for lectures - 15 hrs

- preparation for examination- 15 hrs

Altogether: 60 hrs (2 ECTS)

Student

W1: is familiar with different data segmentation and visualization methods – K_W08

W2: thorough knowledge of quantitative data analysis – in such fields as modeling, data reduction, data visualization – K_W02, K_W05, K_W08

W3: is well acquainted with the concept of supervised and unsupervised learning approaches used in statistics – K_W07, K_W08

W4: knows how to perform advanced analysis of relationships between studied variables using partial least squares regression and correlation method – K_W06, K_W08

Student

U1: is able to design an advanced statistical data analysis pipeline in specialized software (R / SPSS) – K_U01, K_U06, K_U12

U2: can preprocess, classify, and visualize data using supervised and unsupervised algorithm (k-means, hierarchical cluster analysis, decision trees) - K_U01 , K_U09, K_U12

U3: can conduct an analysis of relationship between variables using multivariate correlational and regression approaches (e.g., partial least square correlation / regression) – K_U04, K_U07, K_U12

U4: can perform modeling using SPSS / R software - K_U02, K_U03, K_U12

Student

K1: understands the need for appropriate data preprocessing and visualization – K_K01

K2: is well prepared to critically analyze data and draw well-informed conclusions – K_K02

K3: is aware of power and limitations of statistical tools – K_K02

K4: while understanding own limitations and strong points, is able to add value to team works in advanced analytical projects - K_K05

Expository teaching methods:

- informative (conventional) lecture

- informative (conventional) lecture

The course will get students acquainted with selected advanced concepts and methods in statistical data analysis. Students will solidify their knowledge about supervised and unsupervised approaches in data segmentation, visualization, and prediction. Students will learn one of the sophisticated methods used in correlational analysis and prediction.

Advanced statistics is a course that delves deeper into the field of statistics. It is designed to provide students with a comprehensive understanding of some of more advanced analysis methods commonly used in neuroscience.

First, students will be presented with theoretical and practical knowledge on how to use supervised and unsupervised methods used in data preprocessing, visualization, classification, and prediction (e.g., k-means, hierarchical cluster analysis, decision trees).

Then, the course, building on basics in probability theory, statistical inference, and modeling, will introduce advanced approaches for data modeling, such as multivariate partial least squares regression and correlation techniques (PLSR/PLSC).

Students will be prepared to use appropriate tools offered by statistical software packages such as R, SPSS, and other accessible tools for data manipulation in order to design a desired analysis pipeline.

Throughout the course, students are expected to develop a deep understanding of how to interpret statistical results, critically evaluate research findings, and effectively communicate statistical information to others.

List of topics:

1. Unsupervised data reduction, segmentation, and visualization:

– k-means,

- hierarchical cluster analysis

2. Supervised advanced data predictive models

- decision trees

3. Multivariate correlation / regression models

– partial least square correlation

- partial least square regression

4. Writing scripts for advanced data statistical analysis using R / SPSS software

5. Statistical inference using learned methods and tools.

• A. Field, J. Miles, and Z. Field, Discovering Statistics Using R, 1st edition. London ; Thousand Oaks, Calif: SAGE Publications Ltd, 2012.

• A. Field, An Adventure in Statistics: The Reality Enigma, 1st edition. Los Angeles: SAGE Publications Ltd, 2016.

• A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial Least Squares (PLS) methods for neuroimaging: A tutorial and review,” NeuroImage, vol. 56, no. 2, pp. 455–475, May 2011, doi: 10.1016/j.neuroimage.2010.07.034.

• Bahnsen A.C., Aouada D., i Ottersten B., Example-dependent cost-sensitive decision trees, „Expert Systems with Applications”, 2015, t.42, nr 19, s. 6609–6619.

• Kim J.K., Song H.S., Kim T.S., i Kim H.K., Detecting the change of customer behavior based on decision tree analysis, „Expert Systems”, 2005, t.22, nr 4, s. 193–205.

• D. T. Larose, Discovering Knowledge in Data: An Introduction to Data Mining: 4. Hoboken: Wiley, 2014.

• M. Kraska-Miller, Nonparametric Statistics for Social and Behavioral Sciences. Boca Raton, 2013.

• Peng R.D., R Programming for Data Science, 2020.

• R. S. Witte and J. S. Witte, Statistics, 11th Edition. Wiley, 2017.

Assessment methods:

- lecture: oral examination

Assessment criteria:

lecture: oral exam, with three open-ended questions regarding data visualization, modeling, and inference, respectively. Grading:

satisfactory – 1 correct answers

good – 2 correct answers

very good – 3 correct answers