This paper proposes the use of an access point (AP) selection scheme to improve the total uplink (UL) spectral efficiency (SE) of a radio stripe (RS) network. This scheme optimizes the allocation matrix between the total number of APs' antennas and users' equipment (UEs) while considering two state-of-the-art and two newly proposed equalization approaches: centralized maximum ratio combining (CMRC), centralized optimal sequence linear processing (COSLP), sequential MRC (SMRC), and parallel MRC (PMRC). The optimization problem is solved through a low-complexity and adaptive genetic algorithm (GA) which aims to output an efficient solution for the AP-UE association matrix. We evaluate the proposed schemes in several network scenarios in terms of SE performance, convergence speed, computational complexity, and fronthaul signalling capacity requirements. The COSLP exhibits the best SE performance at the expense of high computational complexity and fronthaul signalling. The SMRC and PMRC are efficient solutions alternatives to the CMRC, improving its computational complexity and convergence speed. Additionally, we assess the adaptability of the MRC schemes for two different instances of network change: when a new randomly located UE must connect to the RS network and when a random UE is removed from it. We have found that in some cases, by reusing the allocation matrix from the original instance as an initial solution, the SMRC and/or the PMRC can significantly boost the optimization performance of the GA-based AP selection scheme.