TY - JOUR
T1 - A novel methodological approach for analyzing the multifaceted phenomenon of residential segregation
T2 - The case of Berlin
AU - Masías H., Víctor H.
AU - Stier, Julia
AU - Navarro R., Pilar
AU - Valle, Mauricio A.
N1 - Publisher Copyright:
© 2023
PY - 2023/10/1
Y1 - 2023/10/1
N2 - This paper aims to study the phenomenon of residential segregation in Berlin using a novel methodological approach. Generally, the different dimensions of residential segregation (i.e. driven by ethnic, age-group, gender, and socioeconomic dimension) are explored separately. In this context, we propose a methodological approach based on multivariate statistics to explore and visualize residential segregation in the residential areas of Berlin separately as well as jointly. For this purpose, we design a sequential mixed approach including the following analytical techniques: a) kernel density estimation to estimate the density of sub-populations; b) non-negative matrix factorization to cluster residential areas and provide an intuitive interpretation of the important variables defining the clusters. The results are first presented separately according to migration background, age group, gender, as well as socio-economic variables. Then all the dimensions are analyzed together to have a panoramic view of the phenomenon in the city of Berlin. The results are discussed and compared with contemporary literature related to the topic of residential segregation in Berlin. The advantages and limitations, and future research possibilities based on the proposed approach are also discussed.
AB - This paper aims to study the phenomenon of residential segregation in Berlin using a novel methodological approach. Generally, the different dimensions of residential segregation (i.e. driven by ethnic, age-group, gender, and socioeconomic dimension) are explored separately. In this context, we propose a methodological approach based on multivariate statistics to explore and visualize residential segregation in the residential areas of Berlin separately as well as jointly. For this purpose, we design a sequential mixed approach including the following analytical techniques: a) kernel density estimation to estimate the density of sub-populations; b) non-negative matrix factorization to cluster residential areas and provide an intuitive interpretation of the important variables defining the clusters. The results are first presented separately according to migration background, age group, gender, as well as socio-economic variables. Then all the dimensions are analyzed together to have a panoramic view of the phenomenon in the city of Berlin. The results are discussed and compared with contemporary literature related to the topic of residential segregation in Berlin. The advantages and limitations, and future research possibilities based on the proposed approach are also discussed.
KW - Berlin
KW - Kernel density estimation
KW - Multivariate data analysis
KW - Non–negative matrix factorization
KW - Residential segregation
KW - Unsupervised machine learning
UR - http://www.scopus.com/inward/record.url?scp=85165099350&partnerID=8YFLogxK
U2 - 10.1016/j.cities.2023.104465
DO - 10.1016/j.cities.2023.104465
M3 - Artículo
AN - SCOPUS:85165099350
SN - 0264-2751
VL - 141
JO - Cities
JF - Cities
M1 - 104465
ER -