Assignments WS 2019/2020

Simulation Proposal - Shopping Centre Management

Each store in a shopping centre pays rent to the shopping centre landlord. Final rent that a store has to pay is calculated as a sum of fixed rent and percentage rent (based on a percentage of store sales). A company that manages shopping centres is profitable when all tenants of the centre are profitable. Sales among tenants are distributed by normal distribution with one outlier, which has double the sales of the second best. This tenant is typical for each shopping centre and is called an anchor tenant. Usually it is a hypermarket. The managing company has services costs, maintenance costs, marketing costs, which are distributed among the tenants in addition to rent. If a tenant can't afford to pay rent + costs, he will terminate lease and managing company loses profit from rent. Other variables affecting the profit are for example footfall of the centre (number of visitors), number of stores, size of the store, conversion ratio (ratio of people walking in to people buying something), average basket (average money spent per customer), anchor tenant sales ratio to all sales, average length of lease, percentage rent, risk of a tenant going out of business. Since getting sales data of each tenant in a shopping centre is not possible (confidential information), I will build the model with a reasonable deal of abstraction on freely available financial data and retail reports about shopping centers. I will study what percentage rent and fixed rent per square meter should the managing company employ to increase the probability of staying profitable in long term.

Title: Shopping Centre Management

Author: Dominik Tkáč, TKAD01 (talk) 00:00, 13 December 2019 (CET)

Model type: Stock and flow diagram

Modeling tool: Vensim

Oleg.Svatos (talk) 22:03, 13 December 2019 (CET) Approved. Make sure that you describe in detail in the report, how you have derived the equations from the available data.

Simulation Proposal - Orders completion in a warehouse

We need to optimize number of warehouse workers and packaging stations in a warehouse with cca 20 000 different products stored. There are workers that work on picking the products for orders from stock positions and there is also one worker per packaging station assigned. Picking products for one regular order takes 15 minutes on average and for one B2B order takes 90 minutes on average. Packaging one regular order takes 2 minutes on average and B2B order takes 45 minutes on average. Ratio of regular to B2B orders is 25:1. We want to find out optimal set up of number of workers and packaging stations.

Title: Orders completion in a warehouse

Author: Ján Káva Xkavj12 (talk) 12:30, 11 December 2019 (CET)

Model type: Discrete event

Modeling tool: Simprocess

For Simprocess, we require a simulation of a real company / problem. Where you get your data from? Tomáš (talk) 15:46, 14 December 2019 (CET)

The proposal is based on a real warehouse located in Ostrava for which I provide IT services, I have access to warehouse ERP data as well as I can consult any estimations with logisitics specialists. Xkavj12 (talk) 22:08, 14 December 2019 (CET)

Simulation Proposal – Cotton Processing Quality: Cotton, Seed, Lint, Trash

This simulation is inspired by previous experience as process management specialist in Cotton Processing Company in Azerbaijan. According to process, cotton was received from farmers and price was indicated based on the moisture and trash level of the raw material. Then raw cotton is going through individual steps of processing. Different types of products are being manufactured from the raw cotton with different qualities and prices. The purpose of simulation is to analyze possible income of the company and create decision making threshold regarding acceptance and price of raw cotton received from farmers.

Title: Cotton Processing Quality: Cotton, Seed, Lint, Trash

Author: Ibrahim Aghazada

Model type: Stock and flow diagram

Modeling tool: Vensim

Simulation Proposal - Atlantic Bluefin Tuna Population

The population of tuna worldwide is decreasing due to being overfished. Yet some studies claim, that overfishing is not the main cause of the tuna population reduction. Christelle Ravier and Jean-Marc Fromentin in their paper "Long-term fluctuations in the eastern Atlantic and Mediterranean bluefin tuna population" came to the conclusion that the bluefin tuna population may be influence by biotic and environmental factors more than by overexploitation. This simulation will use available data of population size, juvenile death, environmental influences, fishing statistics and other to investigate whether overfishing is the main culprit for bluefin tuna population decline.

Title: Atlantic Bluefin Tuna Population

Author: Petra Vokálová, Vokp00 (talk) 07:43, 9 December 2019 (CET)

Model type: Stock and flow diagram

Modeling tool: Vensim

Oleg.Svatos (talk) 22:09, 11 December 2019 (CET) Approved. Make sure that you describe in detail in the report, how you have derived the equations from the available data (proving so that they are based on relevant data).

Simulation Proposal - Black Friday Walmart Simulation

Various items in Walmart store have different values for customers and therefore are sought out by more agents. Agents (customers) decide whether to try to get the item based on the number of people already on the way to the item and also based on distance from the item. Agents try to get as much value in items, pay for them and get out before the store closes.

Title: Black Friday Walmart Simulation

Author: VÁclav Pleskač, plev00, 13 December 2019

Model type: Multi Agent

Modeling tool: Netlogo

I don't understand the point of the simulation. What is it good for? How you verify it provides meaningul results? How do you measure the value of the product for a customer? Tomáš (talk) 15:50, 14 December 2019 (CET)

Simulation Proposal - Crop Rotation in sustainable farming

Crop rotation is based on growing a series of different types of crops in the same area in sequential seasons. The planned rotation may vary from a growing season to a few years or even longer periods. It is one of the most effective agricultural control strategies that is used in preventing the loss of soil fertility. It also helps in reducing soil erosion and increases crop yield. Planning an effective crop rotation requires weighing fixed and fluctuating production circumstances: market, farm size, labor supply, climate, soil type, growing practices, etc. In this simulation I will try to find parameters which have impact on the whole process of crop rotation with goal to find model providing desired outputs - high soil fertility, low soil erosion, good quality crops, high crop field and therefore higher revenue.

Title: Crop Rotation in sustainable farming

Author: Michal Šimánek, Simm04 (talk) 20:33, 10 December 2019 (CET)

Model type: Stock and flow diagram

Modeling tool: Vensim

Oleg.Svatos (talk) 22:13, 11 December 2019 (CET) What data will you build (derive the equations) your simulation on?

Simm04 (talk) 20:02, 12 December 2019 (CET) I would use available dataset from https://www.nature.com/articles/s41598-017-14271-6#Sec16

Oleg.Svatos (talk) 22:00, 12 December 2019 (CET) OK. Approved. Make sure that you describe in detail in the report, how you have derived the equations from the available data (proving so that they are based on relevant data).

Simulation Proposal - Time Series Model Building Process

Linear dynamic models, including univariate case Autoregressive Integrated Moving Average (ARIMA) models and multivariate case Vector AutoRegression (VAR) models, are mainstream approaches to modeling and forecasting time series in econometric practice. They are popular mainly for their simplicity. However, the model building process is not completely straightforward, and the practitioner has to make some decisions such as the transformation of time series or selecting a number of lags (orders p and q in ARIMA model). Recommendations for the decisions are usually formulated as 'cookbooks.' Since the different sources recommend different decision rules, the cookbooks may provide contradictory results. I propose to study decision rules for orders of autoregressive univariate models using MC simulation. Specifically, I will create a set of data-generating processes, simulate the time series, build the models using the cookbooks, and compare the success rate of building a correct model. Based on the success rates of decision rules, I will compare the relative performance of different rules. Additionally, I will use different data-generating processes to study performance under varied assumptions. The choice rules will include General-To-Specific (GETS), specific-to-general, Schwarz-Bayesian Information Criterion (SBIC), and Akaike Information Criterion (AIC).

Title: Simulation Proposal – Time Series Model Building Process

Author: Nikola Hažmuková, Hazn00 (talk) 18:59, 13 December 2019 (CET)

Model type: Monte Carlo

Modeling tool: Python

This is on the edge of our scope, however sounds interesting, hence worth to try. Please, just bear in mind that the solution must be an original work for this course. If you would use any part done for any other occasion (another course, work, etc.), it is necessary to distinguish it clearly in your paper. Approved. Tomáš (talk) 15:59, 14 December 2019 (CET)