I am looking to find out how can I extract maximum information from PCAP files into both single and multiple excel format files for machine learning classification.
The informations depend on what do you want analyze. You may write you own parser (python and c++ are the most use for this work) based on libpcap. For example tcpdump [http://www.tcpdump.org/], or WinDump for windows [http://www.winpcap.org/windump], snort [https://www.snort.org/] , nmap [http://www.nmap.org/] are tools that may help you. Nmap and Snort are ones of the most uses tools for informations security experts and data analysts for extract information of networks traffics and use.
I have used tshark to get the required content using commands that have Fields -q ip.port ect... and later dumped the output into a tsv files and then used python for parsing them, then dumped the result into json format.