Remove redundant switches used in run_constraint_dora_poly example

[mdmitry1]

Now run_constraint_dora_poly has 24 switches.
In order to use this example for quick start, we need to reduce number of switches as much as possible

sed -n '/^smlp/,/log/p' $(git rev-parse --show-toplevel)/tutorial/examples/constraint_dora/smlp/run_constraint_dora_poly | cat -n

     1	smlp -data $csv \
     2	    -spec $script_path/$json \
     3	    -out_dir ./ \
     4	    -pref $csv:r \
     5	    -mode optimize \
     6	    -pareto f \
     7	    -opt_strategy eager \
     8	    -resp Y1 \
     9	    -feat X1,X2 \
    10	    -model poly_sklearn \
    11	    -tree_encoding flat \
    12	    -compress_rules t \
    13	    -mrmr_pred 0 \
    14	    -epsilon 0.0000005 \
    15	    -delta_rel 0 \
    16	    -rad_rel 0 \
    17	    -save_model t \
    18	    -model_name $csv:r_model \
    19	    -save_model_config t \
    20	    -plots f \
    21	    -pred_plots f \
    22	    -resp_plots f \
    23	    -seed 10 \
    24	    -log_time f \

[mdmitry1]

PR #104 resolves this issue forREADME.md in repository root:

smlp_args=(
    -data ${name}.csv.gz   # input CSV dataset
    -spec ${script_path}/${name_lc}.json  # JSON spec file
    -pref ${name}          # output file prefix
    -mode optimize         # operation mode
    -model poly_sklearn    # model type
    -epsilon 0.0000005     # convergence threshold
)

[mdmitry1]

Fixed