Extrapolation beyond a dated point is always dangerous since you don't have more information. If the sedimentation rate within the dated section remains quite stable and the sedimentology doesn't change too much you could very carefully assume a constant sedimentation rate and extrapolate linearly. However, you'll never be certain and it'll likely be picked up by reviewers as well if the age-depth model in the undated section is important for your overall discussion.
Extrapolation is possible in age depth model scripts such as Bacon ( https://chrono.qub.ac.uk/blaauw/bacon.html ) and Clam ( http://www.chrono.qub.ac.uk/blaauw/clam.html ).
Loic Piret has provided you with with a very good answer. Given the potential variation in past rates of sedimentation as well as unconformities due to past erosional events, many geomorphological contexts will not be amenable to extrapolating age/depth estimations for other sedimentary deposits of materials within those sediments. More productive would be concentrating on multiple methods of age determination for different components of any depositional sequence to look at variation in the pace & tempo (to use George Gaylord Simpson's more elegant term than Stephen J. Gould's use of "punctuated equilibrium") of accretional events. Local or regional landscape reconstruction, or chronological framing of a particular location, will benefit more from careful use of multiple methodologies than seeking shortcuts that, as Loic mentions, will only result in yet more untested assumptions.