Analiza međusobne povezanosti prikupljenih senzorskih podataka

undergraduate thesis

• Title (english): Analysis of the interconnection of collected sensor data
• Author: Ana Skukan
• Mentor: Gordan Ježić (mentor)
• Committee member: Gordan Ježić (predsjednik povjerenstva)
• Committee member: Katarina Mandarić (član povjerenstva)
• Committee member: Ivan Kralj (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-09-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U današnje doba analize podataka i manipulacija istih igra veliku ulogu u poboljšanju kvalitete života. Računalna znanost i njezina grana znanost o podacima su trenutno tražene pozicije za sve aspekte života. Iz tog razloga se poljoprivreda, vrlo bitna grana ljudskog života, razvija u smjeru analize podataka radi maksimizacije koristi koje dobivaju od farmi kao i radi same kvalitete proizvoda. U poljoprivrednim analizama bitan faktor je kredibilnost izvora od kojeg dobivaju podatke kao i... More analize istoga.

Kroz ovaj rad bili su analizirani podaci različitih izvora kako bi se izračunala njihova korelacija. Za izvore su bili korišteni Agri4Cast i Copernicus, a korelacija se računala po svakome mjesecu 2019. godine za varijable temperature, padalina, sunčeve radijacije i vjetra. Cilj je bio time zaključiti koreliraju li izvori i u kojoj mjeri. Mjerenja su bila prikazana raspršenim grafovima gdje je na x-osi uvijek bio datum u mjesecu, a na y-osi varijabla za koju se radi analiza. Za korelacije su bili korišteni Pearsonov, Spearmanov i Kendallov koeficijent, a izračuni su se prikazivali tabličnim i grafičkim putem. Koeficijentima se računala jačina korelacije, ovisno o tome ponašaju li se varijable linearno ili ne te kolika je razlika između istih.

Kako bi mjerenja i vizualizacije bili efikasni i brzi, koristio se Python te pripadajuće knjižice za analize podataka (NumPy, Matplotlib, Pandas i Scipy) koje imaju vrlo jednostavnu vizualizaciju i reprezentaciju podataka. Kroz analize u radu zaključeno je da postoje jake korelacije određenih varijabli no i disbalans u drugim varijablama. Od 4 varijable koje su se koristile u izračunima, dvije imaju jaku korelaciju, jedna ima srednje jaku, a jedna ima slabu do srednje jaku korelaciju (uz iznimke jake korelacije u nekim mjesecima). Prema dobivenim podacima je jasno da izvori nisu kompletno kompetentni za sve podatke, a zaključak je da koreliraju no variraju od slabe do jake korelacije sukladno analiziranim varijablama. Less

Abstract (english)

In today's day and age data analysis and data manipulation play a big part in improving the quality of life. Computer science and its branch data science are one of the most wanted positions for all aspects of life. For that reason, agriculture started developing in the direction of data analysis for maximization of usage of farms and for product quality. In agriculture analysis very important factor is credibility of source from which data comes and credibility of analysis of data.... More

Through this thesis data from different sources was analysed so that correlation between them could be calculated. For sources were used Agri4Cast and Copernicus and correlation was calculated for each month of 2019. for variables of temperature, precipitation, radiation, and wind. The goal was to conclude do those two sources correlate and how much. Source data was interpreted through scatter graphs where x axis had dates of the month, and y axis had variable that was being analysed. For correlations were used three different coefficients: Pearson’s, Spearman’s, and Kendall’s. Calculations were presented tabularly and graphically. With coefficients was calculated how strong the data correlation is and do variables act linearly.

Python was used with proper libraries for data analysis (NumPy, Matplotlib, Pandas and Scipy) so that visualization and calculations would go faster and more efficient. Through analysis in this paper, it was concluded that strong correlations exist with some variables but there is lack of correlation on others. Out of 4 variables that were used in calculations, two have strong correlation, one has medium correlation, and one has weak to medium correlation (with exceptions of few months). By given data sources do not match completely for all data and conclusion is that they correlate but in range from weak to strong correlation, depending on the data.
Less

• Keywords: Python; Analize; Podaci; Agrikultura; Meteorologija; Koeficijent; Korelacija; Graf
• Keywords (english): Python; Analysis; Data; Agriculture; Meteorology; Coefficient; Correlation; Graph
• Language: croatian
• URN:NBN: urn:nbn:hr:168:858495
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-14 21:38:35